An expansion method and system for a sodium-ion battery uninterruptible power supply

Abstract

The present application provides an expansion method and system for a sodium-ion battery uninterruptible power supply. The method includes: obtaining the attribute parameters of sodium-ion battery modules in different historical batches, establishing a sodium-ion battery module attribute database, determining adjustment coefficients for the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules, where the attribute parameters include capacity, internal resistance, and voltage; if there is an energy imbalance, query the sodium-ion battery module attribute database to obtain the target charging current under the current state of charge, calculate the required charging time based on the battery capacity and target charging current of the sodium-ion battery module, and control the output of the charger in real time; after charging is completed, re-estimate the capacity, voltage, and internal resistance of sodium-ion battery modules in different batches in the UPS system, and analyze the performance differences and attenuation laws between sodium-ion battery modules in different batches to update the sodium-ion battery module attribute database.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to an expansion method and system for a sodium ion battery uninterruptible power supply. Background Art

[0002] In the Uninterruptible Power System (UPS) system, there are deviations in the capacity, internal resistance and voltage of different batches of battery modules, which will cause an imbalance in the energy distribution of the system. This imbalance will be more obvious when adding or removing battery modules. How to dynamically adjust the charging current and charging time according to the characteristics of different batches of battery modules to achieve energy balance of the UPS system after adding or removing battery modules is a technical problem that needs to be solved urgently. First, the capacity, internal resistance and voltage deviations of different batches of battery modules will cause them to exhibit different characteristics during the charging and discharging process. This difference makes the charging state and health status of each battery module in the UPS system inconsistent, causing an imbalance in the energy distribution of the system. Secondly, when adding or removing battery modules, the difference between the newly added or removed battery modules and the original battery modules will further aggravate the energy imbalance of the system. This imbalance not only affects the performance and reliability of the UPS system, but also may accelerate the aging and loss of the battery modules. Therefore, how to monitor and dynamically adjust the charging current and charging time according to the characteristics of different batches of battery modules in real time to achieve energy balance of the UPS system after adding or removing battery modules is a complex technical challenge. Summary of the invention

[0003] The present invention provides a method for expanding a sodium ion battery uninterruptible power supply, which mainly comprises:

[0004] Obtain the property parameters of different batches of sodium-ion battery modules in history, establish a sodium-ion battery module property database, and determine the adjustment coefficients of the initial charging current and charging time based on the differences in batches and property parameters of the sodium-ion battery modules. The property parameters include capacity, internal resistance and voltage.

[0005] Real-time monitoring of the property parameters of the battery pack in the UPS system and the output power and remaining time of the UPS system. When the addition or removal of the sodium-ion battery module is detected, it is determined whether there is energy imbalance in the UPS system based on the batch and property parameters of the added or removed sodium-ion battery module, as well as the determined adjustment coefficients of the initial charging current and charging time. The energy imbalance includes capacity deviation, internal resistance deviation and voltage deviation.

[0006] If there is an energy imbalance, the sodium-ion battery module attribute database is queried to obtain the target charging current at the current state of charge. Based on the battery capacity and target charging current of the sodium-ion battery module, the required charging time is calculated to control the charger output in real time.

[0007] During the charging process, continuously monitor the charging status of sodium-ion battery modules in different batches in the UPS system and the system energy change. Based on the differences in attribute parameters among sodium-ion battery modules in different batches, compare several different attribute parameters of the sodium-ion battery modules with the corresponding preset target values respectively. If the attribute parameters of the battery reach the preset comparison result, adjust the charging time and charging current accordingly according to the comparison result;

[0008] After charging is completed, re-estimate the capacity, voltage and internal resistance of sodium-ion battery modules in different batches in the UPS system, and analyze the performance differences and attenuation laws among sodium-ion battery modules in different batches to update the attribute database of sodium-ion battery modules;

[0009] Based on the updated attribute database of sodium-ion battery modules, monitor the UPS system load in real time, analyze the load characteristics including load type, load nature and load change law, record the load peak within a preset time, predict the load growth trend of the UPS system, and calculate the system backup time and capacity redundancy based on the current load, peak load and predicted load growth trend to obtain the extended capacity of the sodium-ion battery modules in the UPS system.

[0010] Optionally, obtaining the attribute parameters of historical sodium-ion battery modules in different batches, establishing an attribute database of sodium-ion battery modules, and determining the adjustment coefficients of the initial charging current and charging time according to the batch and attribute parameter differences of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance and voltage, including:

[0011] Obtain the production batch identification, production time identification and battery parameters of sodium-ion batteries through a data collector, and establish a multi-dimensional attribute data structure according to the recursive partitioning algorithm;

[0012] Use the random forest regression algorithm to analyze the battery capacity parameters, internal resistance values and voltage parameters in the multi-dimensional attribute data structure to obtain the mapping relationship between the charging current value and the charging time length of the battery module;

[0013] Read the battery module power percentage, discharge duration and discharge power from the multi-dimensional attribute data structure according to the mapping relationship, and calculate the charging time reference value;

[0014] Calculate the charging time correction coefficient based on the internal resistance value and voltage parameter of the battery.

[0015] Optionally, the attributes of the battery pack in the real-time monitoring UPS system, as well as the output power and remaining time of the UPS system, are monitored. When a sodium-ion battery module is added or removed, based on the batch and attribute parameters of the added or removed sodium-ion battery module, and the determined adjustment coefficients of the initial charging current and charging time, it is determined whether there is energy imbalance in the UPS system. Energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation, including:

[0016] The voltage value, internal resistance value, capacity value, output power value, and remaining operating time of the battery pack are obtained through a voltage data collector, and a time series feature vector of the battery pack parameters is obtained by processing with a recurrent neural network;

[0017] Based on the time series feature vector of the battery pack parameters, the batch number of the battery module, the charging current parameter, and the charging time coefficient read from the sodium-ion battery manager, a state feature matrix of the battery pack is generated by using a convolutional neural network;

[0018] The capacity deviation value, internal resistance deviation value, and voltage deviation value are calculated for the state feature matrix of the battery pack, and it is determined whether the capacity deviation value, internal resistance deviation value, and voltage deviation value exceed the corresponding pre-stored threshold intervals to obtain an imbalance state identification code;

[0019] The imbalance state identification code is classified by using a recursive binary method to generate high, medium, and low level imbalance warning signals. According to the level of the imbalance warning signal, a compensation coefficient is obtained from the charging parameter database, and the battery energy imbalance compensation value is calculated by combining the charging current parameter and the charging time coefficient.

[0020] Optionally, if there is energy imbalance, the attribute database of the sodium-ion battery module is queried to obtain the target charging current under the current state of charge. Based on the battery capacity and the target charging current of the sodium-ion battery module, the required charging time is calculated, and the output of the charger is controlled in real time, including:

[0021] The surface temperature value of the battery is obtained from a temperature sensor and the battery voltage value is obtained from a voltage collector, and the target charging current parameter is obtained based on the surface temperature value of the battery and the battery voltage value;

[0022] The surface temperature value of the battery and the battery voltage value are processed by using a gradient descent algorithm to obtain a charging control feature vector including the temperature change rate and the voltage change rate;

[0023] The remaining battery charge value is recursively least-squares calculated according to the charging control feature vector to obtain a segmented charging current curve;

[0024] If the output power of the segmented charging current curve exceeds the preset threshold interval, the charging current output value is reduced until the battery voltage value reaches the charging cut-off condition.

[0025] Optionally, during the charging process, continuously monitor the charging status of sodium-ion battery modules in different batches in the UPS system and the system energy change, and based on the differences in attribute parameters existing between different batches of sodium-ion battery modules, respectively compare several different attribute parameters of the sodium-ion battery modules with corresponding preset target values. If the attribute parameters of the battery reach the preset comparison result, adjust the charging time and charging current accordingly, including:

[0026] Obtain the charging status value, charging power value, and remaining power value of the sodium-ion battery module through the monitoring unit, and the monitoring unit outputs the voltage change value, temperature change value, and internal resistance change value according to the attribute data collector;

[0027] Perform data processing on the voltage change value, temperature change value, and internal resistance change value using a recurrent neural network to obtain battery state characteristic parameters;

[0028] Read the battery batch number, corresponding charging current value, charging time length, and discharge depth value from the charging controller, and use the battery state characteristic parameters to divide the parameter groups of different batches of battery modules;

[0029] Calculate the charging parameter reference interval for the parameter group, compare the charging parameter reference interval with the preset target value in the preset threshold database to obtain a deviation value, and generate a charging current adjustment amount and a charging time adjustment amount according to the deviation value.

[0030] Optionally, after the charging is completed, re-estimate the capacity, voltage, and internal resistance of sodium-ion battery modules in different batches in the UPS system, and analyze the performance differences and attenuation laws between different batches of sodium-ion battery modules to update the sodium-ion battery module attribute database, including:

[0031] Obtain the capacity measurement value, voltage measurement value, and internal resistance measurement value of the battery module in the fully charged state collected by the data collector, and filter the measured values through a moving average filter to obtain smoothed measurement data;

[0032] Calculate the capacity attenuation rate, voltage attenuation rate, and internal resistance growth rate of the battery module according to the smoothed measurement data, and obtain performance deviation data by comparing with the preset reference performance value;

[0033] Process the performance deviation data using a long short-term memory neural network to obtain a battery performance characteristic matrix, calculate a performance comprehensive score according to the battery performance characteristic matrix, and divide a performance level identification code;

[0034] Determine the data update time interval value according to the performance level identification code, and process the smoothed measurement data using the recursive least squares method to obtain predicted values of the capacity, voltage, and internal resistance parameters.

[0035] Optionally, based on the updated sodium-ion battery module attribute database, the UPS system load is monitored in real time, and the load characteristics including load type, load nature, and load change law are analyzed, and the load peak value within a preset time is recorded, and the load growth trend of the UPS system is predicted. Based on the current load, peak load, and predicted load growth trend, the system backup time and capacity redundancy are calculated to obtain the extended capacity of the sodium-ion battery module of the UPS system, including:

[0036] When obtaining power load data through the load monitoring unit, the load monitoring unit classifies the power load data according to resistive load, inductive load, and capacitive load to obtain load classification data;

[0037] Calculate the load characteristic vector according to the load classification data, extract the peak load amount and peak-valley ratio value from the load characteristic vector to obtain the load power ratio and equipment utilization rate;

[0038] Use the support vector regression method to process the load power ratio and the equipment utilization rate to obtain a load growth curve;

[0039] Calculate the backup power supply time according to the load growth curve, obtain the standard capacity specification from the battery attribute database, and determine the battery capacity expansion value.

[0040] The embodiment of the present invention also provides an extended system for a sodium-ion battery uninterruptible power supply, including:

[0041] A database establishment module for obtaining the attribute parameters of sodium-ion battery modules in different historical batches, establishing a sodium-ion battery module attribute database, and determining the adjustment coefficients of the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage;

[0042] A monitoring and judgment module for monitoring in real time the attribute parameters of the battery pack in the UPS system, as well as the output power and remaining time of the UPS system. When it is detected that a sodium-ion battery module is added or removed, according to the batch and attribute parameters of the added or removed sodium-ion battery module, and the determined adjustment coefficients of the initial charging current and charging time, it is judged whether there is energy imbalance in the UPS system. The energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation;

[0043] A processing and calculation module for querying the sodium-ion battery module attribute database when the monitoring and judgment module determines that there is energy imbalance in the UPS system, obtaining the target charging current under the current state of charge, calculating the required charging time based on the battery capacity and target charging current of the sodium-ion battery module, and controlling the output of the charger in real time;

[0044] The monitoring and judgment module is further configured to continuously monitor the charging status of sodium-ion battery modules in different batches and the system energy change during the charging process in the UPS system, and based on the attribute parameter differences existing between different batches of sodium-ion battery modules, respectively compare several different attribute parameters of the sodium-ion battery modules with corresponding preset target values. If the attribute parameters of the battery reach the preset comparison result, the charging time and charging current are adjusted accordingly according to the comparison result;

[0045] The monitoring and judgment module is further configured to, after the charging is completed, re-estimate the capacity, voltage and internal resistance of sodium-ion battery modules in different batches in the UPS system, and analyze the performance differences and attenuation laws between different batches of sodium-ion battery modules, so as to update the attribute database of the sodium-ion battery modules;

[0046] The monitoring and judgment module is further configured to, based on the updated attribute database of the sodium-ion battery modules, continuously monitor the UPS system load in real time, analyze the load characteristics including load type, load nature and load change law, and record the load peak value within a preset time, predict the load growth trend of the UPS system, and calculate the system backup time and capacity redundancy based on the current load, peak load and predicted load growth trend, so as to obtain the extended capacity of the sodium-ion battery modules in the UPS system.

[0047] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor on an electronic device, the steps of any one of the above methods are implemented.

[0048] The technical solution provided by the embodiment of the present invention may include the following beneficial effects:

[0049] The present invention discloses an expansion method for a sodium-ion battery uninterruptible power supply. The method establishes an attribute database of sodium-ion battery modules, continuously monitors the battery pack parameters and system output in real time, and judges the energy balance state when detecting changes in the battery modules. If there is an energy imbalance, the charging strategy is adjusted according to the database information, and the charging process is controlled in real time. After the charging is completed, the battery performance is re-evaluated and the database is updated. The present invention also predicts the system load growth trend based on the updated database and load analysis, calculates the backup time and capacity redundancy, and determines the extended capacity of the battery modules. This method can effectively manage the performance differences of different batches of sodium-ion battery modules, optimize the charging strategy, improve the reliability and efficiency of the UPS system, and at the same time provide a basis for the system capacity planning, realizing intelligent battery management and system optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a flowchart of an expansion method for a sodium-ion battery uninterruptible power supply according to the present invention.

[0051] Figure 2Schematic diagram of an expansion method for a sodium-ion battery uninterruptible power supply according to the present invention.

[0052] Figure 3 Another schematic diagram of an expansion method for a sodium-ion battery uninterruptible power supply according to the present invention.

[0053] Figure 4 Schematic diagram of the structure of an expansion system for a sodium-ion battery uninterruptible power supply according to the present invention. Detailed implementation manners

[0054] In order to make the objectives, technical solutions and advantages of the present invention clearer, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0055] As Figure 1 shown, the expansion method for a sodium-ion battery uninterruptible power supply in this embodiment may specifically include:

[0056] Step S101: Obtain the attribute parameters of sodium-ion battery modules in different historical batches, establish a sodium-ion battery module attribute database, and determine the adjustment coefficients of the initial charging current and charging time according to the differences in the batch and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage.

[0057] Obtain the production batch identifier, production time identifier, and battery parameters of the sodium-ion battery through a data collector, and establish a multi-dimensional attribute data structure according to the recursive partitioning algorithm; use the random forest regression algorithm to analyze the battery capacity parameter, internal resistance value, and voltage parameter in the multi-dimensional attribute data structure to obtain the mapping relationship between the charging current value and the charging time length of the battery module; calculate the charging time reference value according to the mapping relationship by reading the battery module power percentage, discharge duration, and discharge power from the multi-dimensional attribute data structure; calculate the charging time correction coefficient based on the internal resistance value and voltage parameter of the battery.

[0058] Specifically, the production batch, production time, battery capacity parameters, internal resistance value, voltage parameters, and module production line identification are obtained from the sodium-ion battery production line through a data collector. After standardizing the collected parameters, a multi-dimensional attribute data structure is established according to the recursive partitioning algorithm. For the battery capacity parameters, internal resistance value, and voltage parameters in the multi-dimensional attribute data structure, the random forest regression algorithm is used to calculate the mapping relationship between the charging current value and the charging time length of each batch of battery modules, and a charging parameter prediction model is constructed. The battery module power percentage, discharge duration, and discharge power are read from the multi-dimensional attribute data structure, the charging time reference value is calculated according to the charging parameter prediction model, and the charging current interval is divided according to the battery capacity. The charging time correction coefficient is calculated based on the internal resistance value and voltage parameters of the battery, and the actual charging time is obtained by combining the charging time reference value. The corresponding charging current upper limit value is queried in the multi-dimensional attribute data structure according to the battery module batch number, and the actual charging current parameter is obtained through the product operation of the charging time correction coefficient and the charging current reference value. The charging control instruction sequence is established according to the actual charging time and the actual charging current parameter, and the charging process segmentation point is set according to the battery power percentage. The production process of the sodium-ion battery module includes multiple key parameter acquisition points. The production batch and production time information are obtained by arranging high-precision acquisition equipment on the production line. For core parameters such as battery capacity, internal resistance, and voltage, a standardized processing method is used for normalization, and the value range is controlled between 0 and 1. The sampling frequency of the data collector is set to 200 times per second, and the sampled data is first denoised and then stored in the multi-dimensional attribute data structure. In the recursive partitioning algorithm, the battery module parameters are clustered according to the production batch. The capacity parameter is used as the first-dimensional feature, the internal resistance parameter is used as the second-dimensional feature, and the voltage parameter is used as the third-dimensional feature to construct the feature space. Each battery module forms a unique feature vector in the feature space. For a battery module with a capacity of 180 ampere-hours, an internal resistance of 2.5 milliohms, and a voltage of 3.2 volts, its feature vector is represented as (0.8, 0.6, 0.7) after normalization. The random forest regression algorithm uses 50 decision trees to construct the charging parameter prediction model. The input features include three parameters: battery capacity, internal resistance, and voltage, and the output results are the charging current value and the charging time length. When the battery capacity is 180 ampere-hours, the charging current reference value is set to 90 amperes, and the charging time reference value is 120 minutes. Considering the internal resistance difference of different batches of battery modules, the charging current interval is divided according to the internal resistance value. When the internal resistance is 2.5 milliohms, the charging current upper limit value is 100 amperes. The charging time correction coefficient is related to the battery voltage parameter. When the voltage is 3.2 volts, the correction coefficient is 1.2, and the actual charging time is 144 minutes. The battery module batch number is associated with the charging current upper limit value, and the limit data is obtained by looking up the table. The actual charging current parameter is obtained by multiplying the correction coefficient and the reference value to get 108 amperes.During the charging control process, segmentation points are set according to the battery power. Different charging strategies are adopted when the power reaches 30%, 60%, and 90% respectively. The constant current charging method is used in the 0-30% stage, with a charging current of 108 amperes. The charging current drops to 80 amperes in the 30-60% stage and to 50 amperes in the 60-90% stage. The constant voltage charging method is used in the 90-100% stage.

[0059] Step S102: Monitor the attribute parameters of the battery pack in the UPS system, as well as the output power and remaining time of the UPS system in real time. When it is detected that a sodium-ion battery module is added or removed, judge whether there is energy imbalance in the UPS system according to the batch and attribute parameters of the added or removed sodium-ion battery module, as well as the determined initial charging current and the adjustment coefficient of the charging time. The energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation.

[0060] Obtain the battery pack voltage value, internal resistance value, capacity value, output power value, and remaining operating time through a voltage data collector, and use a recurrent neural network to process to obtain the time series feature vector of the battery pack parameters; according to the time series feature vector of the battery pack parameters and the battery module batch number, charging current parameter, and charging time coefficient read from the sodium-ion battery manager, use a convolutional neural network to generate the battery pack state feature matrix; calculate the capacity deviation value, internal resistance deviation value, and voltage deviation value for the battery pack state feature matrix, and judge whether the capacity deviation value, internal resistance deviation value, and voltage deviation value exceed the corresponding threshold intervals stored in advance to obtain the imbalance state identification code; use the recursive binary method to classify the imbalance state identification code to generate high, medium, and low level imbalance warning signals, and obtain the compensation coefficient from the charging parameter database according to the level of the imbalance warning signal, and calculate the battery energy imbalance compensation value in combination with the charging current parameter and the charging time coefficient.

[0061] Specifically, a voltage data collector is used to obtain voltage values, internal resistance values, capacity values, output power values, and remaining operating time from the battery pack. The collected signals are processed to remove outliers by means of maximum and minimum value clipping. A recurrent neural network is employed to construct a time-series feature vector of the battery pack parameters. The batch number of the battery module, charging current parameters, and charging time coefficient are read from the sodium-ion battery manager. Combining with the time-series feature vector of the battery pack parameters, a convolutional neural network is used to generate a state feature matrix of the battery pack. The voltage threshold, internal resistance threshold, and capacity threshold pre-stored in the battery manager are read, and the capacity deviation value, internal resistance deviation value, and voltage deviation value are calculated based on the state feature matrix of the battery pack. The capacity deviation value, internal resistance deviation value, and voltage deviation value are compared with the corresponding thresholds. If they exceed the threshold range, it is determined as an energy imbalance state, and a corresponding imbalance state identification code is generated. The recursive binary method is used to classify the imbalance state identification code, and high, medium, and low-level imbalance alarm signals are generated accordingly. According to the level of the imbalance alarm signal, the compensation coefficient is extracted from the charging parameter database, and the battery energy imbalance compensation value is calculated by combining the charging current parameter and the charging time coefficient. The monitoring of the battery pack parameters in the UPS system involves multiple key indicators. A high-precision voltage data collector is used to perform real-time sampling on the battery pack, and the sampling frequency is set to 500 times per second to collect voltage values, internal resistance values, capacity values, output power values, and remaining operating time. For the collected original signals, the voltage clipping range is set between 2.8 volts and 3.6 volts, and the internal resistance clipping range is set between 2 milliohms and 5 milliohms. Data outside this range is determined as an outlier and removed. After the sampled data is input into the recurrent neural network, a time-series feature vector is generated, which reflects the dynamic change trend of the battery pack parameters. The basic information such as the production batch and material code of the battery module is stored in the sodium-ion battery manager, and the charging current parameter and the charging time coefficient are also recorded. When the battery pack is in the charging state, the charging current parameter range is between 20 amperes and 100 amperes, and the charging time coefficient is between 0.8 and 1.5. The convolutional neural network receives the time-series feature vector and the charging parameter data and outputs a 16×16-dimensional state feature matrix of the battery pack. Each element of the matrix corresponds to a battery state feature. The voltage threshold preset in the battery manager is ±0.2 volts, the internal resistance threshold is ±0.5 milliohms, and the capacity threshold is ±5%. Taking a certain group of sodium-ion batteries as an example, if the measured voltage value is 3.4 volts and the nominal value is 3.2 volts, the calculated voltage deviation value is 0.2 volts, reaching the upper limit of the preset threshold, and the system determines that there is a voltage imbalance in this group of batteries. Similarly, if the measured internal resistance value is 4.2 milliohms and the nominal value is 3 milliohms, the internal resistance deviation value is 1.2 milliohms, exceeding the threshold range, and it is determined that there is an internal resistance imbalance. For the determined imbalance state, the system uses the recursive binary method for classification. It is set that when the voltage deviation exceeds 50% of the threshold, it is a high-level alarm, when it exceeds 20% to 50%, it is a medium-level alarm, and when it is less than 20%, it is a low-level alarm.When a high-level alarm occurs, extract the compensation coefficient 0.7 from the charging parameter database and multiply it by the original charging current of 80 amperes to obtain a compensated charging current of 56 amperes. The intermediate alarm corresponds to a compensation coefficient of 0.8, and the low-level alarm corresponds to a compensation coefficient of 0.9, realizing the adaptive adjustment of charging parameters.

[0062] In step S103, if there is an energy imbalance, query the sodium-ion battery module attribute database to obtain the target charging current under the current state of charge. Based on the battery capacity of the sodium-ion battery module and the target charging current, calculate the required charging time and control the output of the charger in real time.

[0063] Obtain the battery surface temperature value from the temperature sensor and the battery voltage value from the voltage collector, and obtain the target charging current parameter according to the battery surface temperature value and the battery voltage value; use the gradient descent algorithm to process the battery surface temperature value and the battery voltage value to obtain a charging control feature vector including the temperature change rate and the voltage change rate; perform recursive least squares calculation on the battery remaining power value according to the charging control feature vector to obtain a segmented charging current curve; if the output power of the segmented charging current curve exceeds the preset threshold range, reduce the charging current output value until the battery voltage value reaches the charging cut-off condition.

[0064] Specifically, a battery imbalance identifier of a sodium-ion battery module is read by a real-time data collector, the surface temperature of the battery is obtained from a temperature sensor, and according to the state of charge value and the battery capacity value, the target charging current parameter of the corresponding working range is obtained from an attribute database. The temperature change rate is calculated based on the temperature sensor data, the voltage data is obtained from a voltage collector to calculate the voltage change rate, and a charging control feature vector including the target charging current, the temperature change rate, and the voltage change rate is constructed using the gradient descent algorithm. The initial charging time is calculated through the percentage of remaining power and the capacity attenuation rate, the initial charging time is corrected based on the charging control feature vector, and a segmented charging current curve is generated using the recursive least squares method. The output power of the charger for each time period is calculated according to the segmented charging current curve, and a charging control instruction including the current magnitude and the duration is generated. The actual output current value is read from the charger current detection unit and compared with the charging control instruction. When the deviation exceeds the preset range, the charging current is corrected. Based on the data of the charger power detection unit, it is judged whether the output power exceeds the preset threshold range. When it exceeds, the output value of the charging current is reduced. Through the data of the battery voltage detection unit and the temperature detection unit, it is judged whether the battery reaches the charging cut-off condition. When the cut-off condition is reached, a charging stop instruction is generated. The charging control of sodium-ion batteries involves the real-time monitoring of multiple key parameters. The temperature data is obtained by sampling 10 times per second through the temperature sensors arranged on the battery surface, and the real-time data collector synchronously obtains parameters such as the imbalance identifier, the state of charge value, and the battery capacity value. When the battery pack is in an energy imbalance state, the attribute database returns the corresponding target charging current parameter according to the state of charge. For example, when the state of charge is 80%, the corresponding target charging current is 20 amperes. The temperature change rate reflects the heat generation situation during the battery charging process and is calculated from the temperature sampling data for 60 consecutive seconds. During normal operation, the temperature change rate is within 0.1 degrees Celsius per minute. The voltage data sampling frequency is 100 times per second. The voltage change rate characterizes the voltage response characteristics during the charging process. When the voltage change rate exceeds 0.1 volts per minute, it indicates that the charging current is too large. The charging control feature vector includes three parameters: the target charging current, the temperature change rate, and the voltage change rate. The feature vector is optimized using the gradient descent algorithm. Taking a sodium-ion battery with a capacity of 180 ampere-hours as an example, when the initial state of charge is 30%, the initial charging time calculated according to the remaining power is 4 hours, and the actual charging time is 4.5 hours after correction by combining the temperature and voltage change rates. The segmented charging current curve uses a constant current charging of 60 amperes in the 0-30% interval, 40 amperes in the 30-80% interval, and 20 amperes in the 80-100% interval. The charger current detection unit monitors the output current in real time with a sampling frequency of 1000 times per second. When the deviation between the detected actual current value and the control instruction exceeds 2 amperes, the current is corrected. The charger power detection unit monitors the output power. The preset power threshold is 3000 watts. When it exceeds the threshold, the charging current is gradually reduced in steps of 100 watts.When the battery voltage reaches 3.8 volts or the temperature exceeds 45 degrees Celsius, it is determined that the charging cut-off condition is reached, and a charging stop command is generated to cut off the output of the charger. During the entire charging process, the charging controller continuously performs closed-loop control, and by adjusting the charging current in real time, it keeps the battery temperature and voltage within a safe range. The dynamic adjustment of the charging current takes into account both the state of charge of the battery and the change trends of temperature and voltage, achieving precise control of the battery charging process. When energy imbalance occurs in the battery, the charging controller can correct the imbalance state by adjusting the charging parameters to ensure the charging balance of the battery pack.

[0065] Step S104, during the charging process, continuously monitor the charging status of sodium-ion battery modules in different batches in the UPS system and the energy change of the system. Based on the differences in attribute parameters among different batches of sodium-ion battery modules, respectively compare several different attribute parameters of the sodium-ion battery modules with the corresponding preset target values. If the attribute parameters of the battery reach the preset comparison result, adjust the charging time and charging current accordingly according to the comparison result.

[0066] Obtain the charging status value, charging power value, and remaining power value of the sodium-ion battery module through the monitoring unit. The monitoring unit outputs the voltage change value, temperature change value, and internal resistance change value according to the attribute data collector; perform data processing on the voltage change value, temperature change value, and internal resistance change value using a recursive neural network to obtain the battery state characteristic parameters; read the battery batch number and the corresponding charging current value, charging time length, and discharge depth value from the charging controller, and divide the parameter group for different batches of battery modules using the battery state characteristic parameters; calculate the reference interval of the charging parameters for the parameter group, compare the reference interval of the charging parameters with the preset target value in the preset threshold database to obtain the deviation value, and generate the charging current adjustment amount and charging time adjustment amount according to the deviation value.

[0067] Specifically, the monitoring unit collects the charging status, charging power value, and remaining power value of the sodium-ion battery module in batches. Based on the attribute data collector, the voltage change value, temperature change value, and internal resistance change value are obtained, and the collected data is denoised and smoothed. The denoised data is processed using a recurrent neural network. The input features include the voltage change value, temperature change value, and internal resistance change value, and the battery state characteristic parameters are output. The battery batch number, corresponding charging current value, charging time length, and discharge depth value are obtained from the charging controller. According to the battery state characteristic parameters, the battery modules of different batches are parameter-grouped. Based on the parameter-grouping results, the mean and variance of the charging parameters of each group are calculated to generate a reference interval for the charging parameters. The preset target values of each batch of batteries are read from the preset threshold database and compared with the reference interval for the charging parameters to generate a deviation value. The correction level is divided according to the size of the deviation value, and the charging current adjustment amount or charging time adjustment amount is generated according to the correction level. The adjusted charging parameters are input into the charging controller, and the actual output parameters read from the controller are compared with the target parameters. When the deviation exceeds the threshold, the adjustment amount is recalculated. The acquisition of sodium-ion battery monitoring data involves the real-time acquisition of multiple key indicators. The monitoring unit samples the data of each battery module, and the sampling frequency is set to 1000 times per second. The basic parameters including the charging status, charging power value, and remaining power value are collected. During the acquisition process, the sliding average filtering method is used to denoise the original signal, and the sliding window length is set to 20 sampling points. The smoothed data can reflect the true change trend of the battery parameters. The recurrent neural network receives the processed voltage change value, temperature change value, and internal resistance change value as input features. The network contains 10 hidden layer nodes and outputs the battery state characteristic parameters. Taking a certain batch of sodium-ion batteries as an example, when the voltage change value is 0.5 mV per second, the temperature change value is 0.2 °C per minute, and the internal resistance change value is 0.1 mΩ per hour, the state characteristic parameters output by the recurrent neural network indicate that the battery is in a normal charging state. For different batches of battery modules, the charging controller records parameters such as the charging current value, charging time length, and discharge depth value. For a battery pack with a capacity of 180 Ah, the charging current value is between 20 A and 100 A, the charging time length is between 2 hours and 8 hours, and the discharge depth value is between 10% and 90%. Based on the battery state characteristic parameters, these charging parameters are grouped, and the batteries with the same capacity level are divided into the same group. The mean and variance of the charging parameters are calculated within each group of batteries to generate a reference interval for the parameters. When the mean charging current of a certain group of batteries is 60 A and the variance is 5 A, the reference interval for the charging current is set to 55 A to 65 A. The preset threshold database stores the preset target values of different batches of batteries. The deviation value can be obtained by comparing with the reference interval. The deviation value is divided into three correction levels according to 0-10%, 10%-20%, and more than 20%, corresponding to adjustment amounts of 5%, 10%, and 15% respectively.Taking the charging current adjustment as an example, when the deviation value is 15%, it belongs to the second - level correction, and the charging current needs to be adjusted by 10%. If the original charging current is 60 amperes, the adjusted charging current becomes 54 amperes. The charging controller reads the actual output current every 10 seconds and compares it with the target current of 54 amperes. When the deviation exceeds 2 amperes, a new round of adjustment calculation is triggered.

[0068] Step S105: After charging is completed, re - estimate the capacity, voltage, and internal resistance of different batches of sodium - ion battery modules in the UPS system, and analyze the performance differences and attenuation laws between different batches of sodium - ion battery modules to update the sodium - ion battery module attribute database.

[0069] Obtain the capacity measurement value, voltage measurement value, and internal resistance measurement value of the battery module in the charging - completed state collected by the data collector, and filter the measured values through a moving average filter to obtain smooth measurement data; calculate the capacity decay rate, voltage decay rate, and internal resistance growth rate of the battery module according to the smooth measurement data, and obtain performance deviation data by comparing with the preset reference performance value; use a long - short - term memory neural network to process the performance deviation data to obtain a battery performance feature matrix, calculate a comprehensive performance score according to the battery performance feature matrix and divide the performance level identification code; determine the data update time interval value according to the performance level identification code, and use the recursive least - squares method to process the smooth measurement data to obtain the predicted values of the capacity, voltage, and internal resistance parameters.

[0070] Specifically, the capacity measurement values, voltage measurement values, and internal resistance measurement values of the sodium-ion battery module in the fully charged state are collected in batches by a voltage and current data collector. Historical measurement values are read from the attribute database, and the measurement data is subjected to moving average filtering to generate smoothed measurement data. Based on the smoothed measurement data, the capacity attenuation rate, voltage attenuation rate, and internal resistance growth rate of each batch of battery modules are calculated, and performance deviation data is generated by comparing with the reference performance values in the attribute database. The performance deviation threshold is obtained from the attribute database, the performance deviation data is normalized, and a long short-term memory neural network is used to construct a battery performance feature matrix. The comprehensive performance score is calculated according to the battery performance feature matrix, the performance levels are divided according to the score intervals, and a performance classification identification code is generated. Based on the performance classification identification code, the data update time interval is determined. The high-performance level interval is set long, and the low-performance level interval is set short. The recursive least squares method is used to calculate the change trend of the measurement data, and the predicted values of the capacity, voltage, and internal resistance parameters are generated. The measurement data, performance levels, and parameter predicted values are written into the corresponding batch records in the attribute database to complete the dynamic update of the battery attribute data. The performance evaluation of the sodium-ion battery module involves the collection and analysis of multiple key parameters. The battery in the fully charged state is sampled by a high-precision voltage and current collector, and the sampling frequency is set to 500 times per second, recording three key parameters: capacity, voltage, and internal resistance. The original sampling data is processed by 20-point moving average filtering, and the filtered data can reflect the true change trend of the battery parameters. Taking a sodium-ion battery with a capacity of 180 ampere-hours as an example, the actual capacity measured at the end of charging is 162 ampere-hours, the voltage is 3.2 volts, and the internal resistance is 3.5 milliohms. The reference data when this batch of batteries was put into use is read from the attribute database, with a capacity of 180 ampere-hours, a voltage of 3.6 volts, and an internal resistance of 2.5 milliohms. Through comparison and calculation, the capacity attenuation rate is 10%, the voltage attenuation rate is 11%, and the internal resistance growth rate is 40%. These performance deviation data are normalized and then input into the long short-term memory neural network to construct a 16×16-dimensional performance feature matrix. Each element in the performance feature matrix corresponds to a performance feature, and the comprehensive performance score of 85 points is obtained through weighted calculation. The performance levels are divided according to the score intervals: above 90 points is level one, 80-90 points is level two, 70-80 points is level three, and below 70 points is level four. This battery scores 85 points and is judged to have a second-level performance, generating a performance classification identification code of 0010. The data update time interval changes with the performance level. Level one performance is set to be updated once every 30 days, level two every 20 days, level three every 10 days, and level four every 5 days. The recursive least squares method is used to analyze the trend of the recent measurement data of this batch of batteries, predicting that the capacity will decay to 160 ampere-hours, the voltage will drop to 3.15 volts, and the internal resistance will rise to 3.6 milliohms in the next cycle. These predicted data, together with the current measurement values and the second-level performance identifier, are written into the attribute database to update the performance records of this batch of batteries. Through the dynamic update mechanism, the attribute database always maintains the latest battery performance data, providing data support for optimizing the charging control strategy.The performance grading mechanism dynamically adjusts the update frequency according to the measured data. Batteries with poorer performance are monitored more frequently to promptly detect potential problems. Predictive analysis quantifies the trend of battery performance changes, providing a basis for maintenance decisions.

[0071] Step S106: Based on the updated sodium-ion battery module attribute database, the load of the UPS system is monitored in real time. Analyze the load characteristics including load type, load nature, and load change pattern, and record the load peak within a preset time. Predict the load growth trend of the UPS system. Based on the current load, peak load, and predicted load growth trend, calculate the system's backup time and capacity redundancy to obtain the extended capacity of the sodium-ion battery module of the UPS system.

[0072] When obtaining the power load data through the load monitoring unit, the load monitoring unit classifies the power load data according to resistive load, inductive load, and capacitive load to obtain load classification data; calculate the load feature vector according to the load classification data, extract the peak load amount and peak-valley ratio value from the load feature vector to obtain the load power ratio and equipment utilization rate; use the support vector regression method to process the load power ratio and the equipment utilization rate to obtain the load growth curve; calculate the backup power supply time according to the load growth curve, obtain the standard capacity specification from the battery attribute database, and determine the battery capacity expansion value.

[0073] Specifically, the power load data is collected by the load monitoring unit, recording the load power value and the load change rate. The load is classified according to resistive load, inductive load, and capacitive load, and a recursive neural network is used to generate the load feature vector. The recent load records are read from the load data memory, and the peak load amount and the peak-valley ratio value are extracted based on the load feature vector. The load power ratio and the equipment utilization rate are calculated for different time periods. The support vector regression method is used to process the load data, and the input features include the load power ratio, the equipment utilization rate, and the peak-valley ratio value to predict the future load growth curve. The load value at the key time point is extracted from the load growth curve, and the load growth rate is calculated by combining the current load power value to generate a load prediction report. The rated capacity and the actual capacity of the battery pack are read from the sodium-ion battery attribute database, and the backup power supply time is calculated based on the load prediction report. According to the comparison result between the backup power supply time and the preset time threshold, the required capacity redundancy coefficient is calculated to generate a battery capacity expansion recommended value. According to the difference between the capacity expansion recommended value and the actual capacity, the standard capacity specification is obtained from the battery attribute database to determine the final expanded capacity value. The load monitoring of the UPS system involves the identification and analysis of multiple load types. The power load is sampled in real time by a high-precision load monitoring unit, and the sampling frequency is set to 1000 times per second, recording the load power value and the load change rate. The load classification is based on the power factor and phase angle characteristics. A resistive load is determined when the power factor is greater than 0.95, an inductive load is determined when the power factor is between 0.7 and 0.95 and the current lags the voltage, and a capacitive load is determined when the power factor is between 0.7 and 0.95 and the current leads the voltage. The recursive neural network receives the load sampling data and outputs a 16-dimensional load feature vector. For the load scenario of the data center, the load data memory records the load data of the most recent month. The peak load amount appears between 14:00 and 16:00 on weekdays, with a value of 800 kW, the valley load appears between 2:00 and 4:00 in the early morning, with a value of 200 kW, and the peak-valley ratio value is 4. The equipment utilization rate remains above 75% during the working period and drops to about 30% at night. The load power ratio reflects the load distribution characteristics in different time periods, with the peak period accounting for 40%, the flat period accounting for 45%, and the valley period accounting for 15%. The load prediction uses the support vector regression method, and the training data includes the load records of the past 6 months. Each record contains three features: the load power ratio, the equipment utilization rate, and the peak-valley ratio value. The prediction results show that the peak load will increase to 900 kW in the next month, and the monthly average load growth rate is 12%. The load prediction values at the key time points of 9:00, 14:00, and 20:00 are 600 kW, 850 kW, and 450 kW respectively. The record in the battery attribute database shows that the rated capacity of the sodium-ion battery pack currently configured in the UPS is 1000 kWh, and the actual capacity is 900 kWh. The backup power supply time is calculated based on the load prediction results. When the load is at the peak of 900 kW, the backup power supply time is 1 hour, which is different from the preset threshold of 2 hours.The calculation result of the capacity redundancy factor is 1.8, indicating that the battery capacity needs to be expanded. The standard capacity specification in the battery attribute database is 500 kWh. According to the capacity difference of 800 kWh, it is determined that 2 groups of standard capacity battery modules need to be added. The load characteristic identification provides a basis for capacity planning. By combining historical data to predict the future load growth trend, an accurate capacity expansion plan is finally obtained.

[0074] An embodiment of the present invention also provides an expansion system for a sodium-ion battery uninterruptible power supply, as Figure 4 shown, including the following modules:

[0075] A database establishment module 401, configured to obtain the attribute parameters of sodium-ion battery modules in different historical batches, establish a sodium-ion battery module attribute database, and determine the adjustment coefficients of the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage;

[0076] A monitoring and judgment module 402, configured to continuously monitor the attribute parameters of the battery pack in the UPS system, the output power and remaining time of the UPS system in real time. When it detects the addition or removal of sodium-ion battery modules, it determines whether there is energy imbalance in the UPS system according to the batches and attribute parameters of the added or removed sodium-ion battery modules, and the determined adjustment coefficients of the initial charging current and charging time. The energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation;

[0077] A processing and calculation module 403, configured to query the sodium-ion battery module attribute database when the monitoring and judgment module determines that there is energy imbalance in the UPS system, obtain the target charging current under the current state of charge, calculate the required charging time based on the battery capacity and target charging current of the sodium-ion battery module, and control the output of the charger in real time;

[0078] The monitoring and judgment module 402 is further configured to continuously monitor the charging status of sodium-ion battery modules in different batches and the system energy change in the UPS system during the charging process, and respectively compare several different attribute parameters of the sodium-ion battery modules with corresponding preset target values based on the attribute parameter differences existing between different batches of sodium-ion battery modules. If the attribute parameters of the battery reach the preset comparison result, the charging time and charging current are adjusted accordingly according to the comparison result;

[0079] The monitoring and judgment module 402 is further configured to re-estimate the capacity, voltage, and internal resistance of sodium-ion battery modules in different batches in the UPS system after charging is completed, and analyze the performance differences and attenuation laws between different batches of sodium-ion battery modules to update the sodium-ion battery module attribute database;

[0080] The monitoring and judgment module 402 is further configured to, based on the updated sodium-ion battery module attribute database, monitor the load of the UPS system in real time, analyze the load characteristics including load type, load nature, and load change pattern, record the load peak within a preset time, predict the load growth trend of the UPS system, and calculate the system backup time and capacity redundancy based on the current load, peak load, and predicted load growth trend, so as to obtain the extended capacity of the sodium-ion battery module of the UPS system.

[0081] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor on an electronic device, the steps of any one of the above methods are implemented.

[0082] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. Without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. An expansion method for a sodium-ion battery uninterruptible power supply, characterized in that, The method includes: Obtaining the attribute parameters of sodium-ion battery modules in different historical batches, establishing a sodium-ion battery module attribute database, and determining the adjustment coefficients of the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage. Real-time monitoring the attribute parameters of the battery pack in the UPS system, as well as the output power and remaining time of the UPS system. When it is detected that a sodium-ion battery module is added or removed, determine whether there is an energy imbalance in the UPS system according to the batch and attribute parameters of the added or removed sodium-ion battery module, and the determined adjustment coefficients of the initial charging current and charging time. The energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation. If there is an energy imbalance, query the sodium-ion battery module attribute database to obtain the target charging current at the current state of charge. Based on the battery capacity and the target charging current of the sodium-ion battery module, calculate the required charging time and control the output of the charger in real time. During the charging process, continuously monitor the charging status of sodium-ion battery modules in different batches in the UPS system and the change of system energy. Based on the differences in attribute parameters among sodium-ion battery modules in different batches, compare several different attribute parameters of the sodium-ion battery modules with the corresponding preset target values respectively. If the attribute parameters of the battery reach the preset comparison result, adjust the charging time and charging current accordingly according to the comparison result. After charging is completed, re-estimate the capacity, voltage, and internal resistance of sodium-ion battery modules in different batches in the UPS system, and analyze the performance differences and attenuation laws among sodium-ion battery modules in different batches to update the sodium-ion battery module attribute database. Based on the updated sodium-ion battery module attribute database, real-time monitor the load of the UPS system, analyze the load characteristics. The load characteristics include load type, load nature, and load change law, and record the load peak within a preset time. Based on the current load and the load peak, predict the load growth trend of the UPS system, generate a load growth curve, calculate the backup power supply time according to the load growth curve, combined with the rated capacity and actual capacity of the battery pack, and calculate the capacity redundancy according to the comparison result between the backup power supply time and the preset time threshold to obtain the extended capacity of the sodium-ion battery module of the UPS system.

2. The method according to claim 1, wherein The obtaining the attribute parameters of sodium-ion battery modules in different historical batches, establishing a sodium-ion battery module attribute database, and determining the adjustment coefficients of the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage, includes: Obtaining the production batch identification, production time identification, and battery parameters of the sodium-ion battery through a data collector, and establishing a multi-dimensional attribute data structure according to the recursive partitioning algorithm. Using the random forest regression algorithm to analyze the battery capacity parameters, internal resistance values, and voltage parameters in the multi-dimensional attribute data structure to obtain the mapping relationship between the charging current value and the charging time length of the battery module. Reading the battery module power percentage, discharge duration, and discharge power from the multi-dimensional attribute data structure according to the mapping relationship, and calculating the charging time reference value. Calculate the charging time correction coefficient based on the internal resistance value and voltage parameter of the battery.

3. The method according to claim 1, wherein The method includes: real-time monitoring of the attribute parameters of the battery pack in the UPS system, the output power and remaining time of the UPS system. When it is detected that a sodium-ion battery module is added or removed, based on the batch and attribute parameters of the added or removed sodium-ion battery module, as well as the determined initial charging current and the adjustment coefficient of the charging time, determine whether there is energy imbalance in the UPS system. The energy imbalance includes capacity deviation, internal resistance deviation and voltage deviation. Obtain the battery pack voltage value, internal resistance value, capacity value, output power value and remaining operating time through a voltage data collector, and use a recurrent neural network to process to obtain the time series feature vector of the battery pack parameters. According to the time series feature vector of the battery pack parameters, the battery module batch number, charging current parameter and charging time coefficient read from the sodium-ion battery manager, use a convolutional neural network to generate the battery pack state feature matrix. Calculate the capacity deviation value, internal resistance deviation value and voltage deviation value for the battery pack state feature matrix, and determine whether any one of the capacity deviation value, internal resistance deviation value and voltage deviation value exceeds the corresponding pre-stored threshold interval to obtain the imbalance state identification code. Use the recursive dichotomy method to classify the imbalance state identification code to generate high, medium and low level imbalance warning signals, obtain the compensation coefficient from the charging parameter database according to the level of the imbalance warning signal, and calculate the battery energy imbalance compensation value in combination with the charging current parameter and the charging time coefficient.

4. The method according to claim 1, characterized in that If there is energy imbalance, query the sodium-ion battery module attribute database to obtain the target charging current under the current state of charge. Based on the battery capacity and the target charging current of the sodium-ion battery module, calculate the required charging time and control the output of the charger in real time. The method includes: Obtain the battery surface temperature value from the temperature sensor and the battery voltage value from the voltage collector, and obtain the target charging current parameter according to the battery surface temperature value and the battery voltage value. Use the gradient descent algorithm to process the battery surface temperature value and the battery voltage value to obtain a charging control feature vector including the temperature change rate and the voltage change rate. Perform recursive least squares calculation on the remaining battery charge value according to the charging control feature vector to obtain a segmented charging current curve. If the output power of the segmented charging current curve exceeds the preset threshold interval, reduce the charging current output value until the battery voltage value reaches the charging cut-off condition.

5. The method according to claim 1, characterized in that, During the charging process, continuously monitor the charging state of sodium-ion battery modules in different batches in the UPS system and the energy change of the system. Based on the attribute parameter differences between sodium-ion battery modules in different batches, compare several different attribute parameters of the sodium-ion battery module with the corresponding preset target values respectively. If the attribute parameters of the battery reach the preset comparison result, adjust the charging time and charging current accordingly. The method includes: Obtain the charging state value, charging power value and remaining charge value of the sodium-ion battery module through the monitoring unit. The monitoring unit outputs the voltage change value, temperature change value and internal resistance change value according to the attribute data collector. Perform data processing on the voltage change value, temperature change value, and internal resistance change value using a recurrent neural network to obtain battery state characteristic parameters; Read the battery batch number, corresponding charging current value, charging time length, and discharge depth value from the charging controller, and divide different batch battery modules into parameter groups using the battery state characteristic parameters; Calculate the charging parameter reference interval for the parameter group, compare the charging parameter reference interval with the preset target value in the preset threshold database to obtain a deviation value, and generate a charging current adjustment amount and a charging time adjustment amount according to the deviation value.

6. The method according to claim 1, wherein After the charging is completed, re-estimate the capacity, voltage, and internal resistance of different batch sodium-ion battery modules in the UPS system, and analyze the performance differences and attenuation laws between different batch sodium-ion battery modules to update the sodium-ion battery module attribute database, including: Obtain the capacity measurement value, voltage measurement value, and internal resistance measurement value of the battery module in the charging completed state collected by the data collector, and filter the measured values through a moving average filter to obtain smooth measurement data; Calculate the capacity attenuation rate, voltage attenuation rate, and internal resistance growth rate of the battery module according to the smooth measurement data, and obtain performance deviation data by comparing with the preset reference performance value; Process the performance deviation data using a long short-term memory neural network to obtain a battery performance characteristic matrix, calculate a performance comprehensive score according to the battery performance characteristic matrix, and divide a performance level identification code; Determine the data update time interval value according to the performance level identification code, and process the smooth measurement data using the recursive least squares method to obtain predicted values of capacity, voltage, and internal resistance parameters.

7. The method according to claim 1, characterized in that, Based on the updated sodium-ion battery module attribute database, monitor the UPS system load in real time, analyze the load characteristics, where the load characteristics include load type, load nature, and load change law, record the load peak value within a preset time, predict the load growth trend of the UPS system based on the current load and load peak value, generate a load growth curve, calculate the backup power supply time according to the load growth curve, combined with the rated capacity and actual capacity of the battery pack, and calculate the capacity redundancy according to the comparison result of the backup power supply time and the preset time threshold to obtain the extended capacity of the sodium-ion battery module in the UPS system, including: When obtaining power load data through the load monitoring unit, the load monitoring unit classifies the power load data according to resistive load, inductive load, and capacitive load to obtain load classification data; Calculate the load characteristic vector according to the load classification data, extract the peak load amount and peak-valley ratio value from the load characteristic vector to obtain the load power ratio and equipment utilization rate; Process the load power ratio and the equipment utilization rate using the support vector regression method to obtain a load growth curve; Calculate the backup power supply time according to the load growth curve, obtain the standard capacity specification from the battery attribute database, and determine the battery capacity expansion value.

8. An expansion system for a sodium-ion battery uninterruptible power supply, characterized in that, Including: A database establishment module, configured to obtain the attribute parameters of sodium-ion battery modules in different historical batches, establish a sodium-ion battery module attribute database, and determine the adjustment coefficients of the initial charging current and charging time according to the differences in the batches and attribute parameters of the sodium-ion battery modules. The attribute parameters include capacity, internal resistance, and voltage. A monitoring and judgment module, configured to monitor in real time the attribute parameters of the battery pack in the UPS system, as well as the output power and remaining time of the UPS system. When it detects the addition or removal of a sodium-ion battery module, it determines whether there is an energy imbalance in the UPS system according to the batch and attribute parameters of the added or removed sodium-ion battery module, and the determined adjustment coefficients of the initial charging current and charging time. The energy imbalance includes capacity deviation, internal resistance deviation, and voltage deviation. A processing and calculation module, configured to query the sodium-ion battery module attribute database when the monitoring and judgment module determines that there is an energy imbalance in the UPS system, obtain the target charging current under the current state of charge, calculate the required charging time based on the battery capacity and the target charging current of the sodium-ion battery module, and control the output of the charger in real time. The monitoring and judgment module is further configured to continuously monitor the charging status of sodium-ion battery modules in different batches in the UPS system and the system energy change during the charging process, and compare several different attribute parameters of the sodium-ion battery modules with the corresponding preset target values respectively based on the differences in the attribute parameters among the sodium-ion battery modules in different batches. If the attribute parameters of the battery reach the preset comparison result, the charging time and charging current are adjusted accordingly according to the comparison result. The monitoring and judgment module is further configured to re-estimate the capacity, voltage, and internal resistance of sodium-ion battery modules in different batches in the UPS system after charging is completed, and analyze the performance differences and attenuation laws among the sodium-ion battery modules in different batches to update the sodium-ion battery module attribute database. The monitoring and judgment module is further configured to, based on the updated sodium-ion battery module attribute database, monitor the UPS system load in real time, analyze the load characteristics, where the load characteristics include load type, load nature, and load change law, and record the load peak within a preset time. Based on the current load and the load peak, predict the load growth trend of the UPS system, generate a load growth curve, calculate the backup power supply time according to the load growth curve, combined with the rated capacity and actual capacity of the battery pack, and calculate the capacity redundancy according to the comparison result between the backup power supply time and the preset time threshold to obtain the extended capacity of the sodium-ion battery module of the UPS system.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor on an electronic device, it implements the steps of the method according to any one of claims 1-7.

Images