An optimization method and system for a distributed battery management system

Abstract

This invention provides an optimization method and system for a distributed battery management system. The method includes: collecting cell parameters, scenario parameters, and historical charge / discharge data of the battery cells and evaluating the battery aging stage; invoking a matching cell state calculation algorithm based on the scenario parameters and battery aging stage; calculating the cell state information of the battery cells based on the cell parameters, scenario parameters, and cell state calculation algorithm; dynamically adjusting the balancing threshold and balancing rate based on the cell state information, scenario parameters, and battery aging stage, and executing a balancing strategy to obtain balancing state information; periodically uploading the running data to a cloud server and receiving algorithm update packages to update the adaptive algorithm library; and calculating the overall state information of the battery pack based on the cell state information and balancing state information. This invention can effectively improve the calculation accuracy of cell state, extend battery cycle life, and improve battery safety.

Description

Technical Field

[0001] This invention relates to the field of battery management systems, and more specifically to an optimization method and system for a distributed battery management system. Background Technology

[0002] The Battery Management System (BMS) is a core component that ensures the safe and efficient operation of energy storage devices. Its core function is to achieve intelligent management and maintenance of battery cells, accurately monitor battery status, such as State of Charge (SOC), State of Power (SOP), and State of Energy (SOE), prevent safety hazards such as overcharging, over-discharging, and thermal runaway, extend battery cycle life, and ensure the stable and reliable operation of the battery system.

[0003] In applications with a large number of dispersed batteries (such as electric vehicles and large-scale energy storage systems), distributed BMS architectures have become mainstream due to their significant advantages, including short wiring harness lengths, small PCB areas, distributed computational loads, and strong scalability. The core design concept of distributed BMS is to decompose functions into multiple battery management slave controllers, with the main battery management controller only responsible for the overall state of the battery pack and the coordinated control of the entire vehicle's energy storage system. However, existing distributed BMS technologies still cannot meet the high-precision, high-reliability battery management requirements of complex scenarios, specifically as follows:

[0004] There are issues with fixed algorithms and poor adaptability. Current battery management algorithms for calculating cell state from the controller are mostly fixed configurations, failing to consider the diversity of battery usage scenarios and the dynamic changes during battery aging. On one hand, actual battery usage scenarios are complex and varied. For example, electric vehicles may face low-temperature environments below -30℃, high-temperature environments above 50℃, and high-altitude, low-pressure environments above 4000 meters. The charge-discharge characteristics, internal resistance variations, and thermal stability of cells differ significantly under different scenarios. On the other hand, batteries go through three stages throughout their lifespan: new battery, semi-aged, and near-discard. Core characteristics such as capacity, internal resistance, and cycle stability continuously change at each stage. Fixed algorithms cannot dynamically match these scenario changes and aging evolution, leading to a continuous decline in the accuracy of cell state calculations.

[0005] There is a lack of targeted equalization strategies. Existing equalization control logic operates based on fixed thresholds and rates, without dynamically adjusting to scenario parameters and battery aging conditions. For example, new battery cells have good consistency, but using a lenient equalization threshold can lead to a gradual deterioration in consistency, while using an excessively high equalization rate can cause unnecessary capacity loss. Aging battery cells have poor consistency, and a fixed low threshold equalization can lead to over-equalization, exacerbating cell damage. In low-temperature environments, an excessively fast equalization rate can easily cause lithium plating, while in high-temperature environments, an inappropriate equalization rate may induce thermal runaway. In high-altitude, low-pressure environments, a fixed rate equalization can reduce energy utilization efficiency. These problems make it difficult to maintain battery module consistency, shorten cycle life, and increase safety risks.

[0006] It lacks self-learning and iterative capabilities. Once the algorithm parameters of existing technologies are fixed, they cannot be optimized and upgraded based on actual usage data. As battery usage time increases and the number of scenarios accumulates, the mismatch between the algorithm and actual operating conditions will gradually widen, leading to a continuous decline in battery management performance. Manual intervention is required to update the algorithm, resulting in high maintenance costs and the inability to achieve performance self-optimization throughout the entire life cycle.

[0007] The lack of scene awareness and aging assessment is a significant issue. Existing technologies cannot accurately capture dynamic changes in environmental parameters such as air pressure and humidity, nor can they systematically assess battery aging stages. This results in a lack of data support for algorithm calls and equalization strategy formulation, further exacerbating the problem of poor adaptability.

[0008] Based on the shortcomings of the existing technology, there is an urgent need for an optimization method and system for distributed battery management systems. Summary of the Invention

[0009] Based on the technical problems described above, this invention provides an optimization method and system for a distributed battery management system. By adding a scene perception and aging state assessment module and an adaptive algorithm library, it achieves accurate collection of scene parameters, dynamic assessment of battery aging stages, adaptive algorithm invocation, dynamic adjustment of balancing threshold and rate, and algorithm self-learning iteration. This solves technical problems such as poor adaptability of fixed algorithms, lack of specificity in balancing strategies, and lack of self-learning ability. Ultimately, it improves the accuracy of cell state calculation, extends battery cycle life, enhances safety and stability, and achieves autonomous optimization throughout the entire life cycle.

[0010] Specifically, according to one aspect of the present invention, an optimization method for a distributed battery management system is provided, the method comprising the following steps:

[0011] Step S1, Data Acquisition and Status Assessment: The battery management system acquires the cell parameters of the battery cells in real time from the controller, acquires the scene parameters of the battery cells in real time through environmental sensors, and reads the historical charge and discharge data of the battery cells from the storage unit through the aging feature extraction unit and assesses the aging stage of the battery based on the historical charge and discharge data.

[0012] Step S2, Dynamic Algorithm Invocation: The battery management controller dynamically invokes a matching cell state calculation algorithm from the adaptive algorithm library based on the scenario parameters and the aging stage of the battery.

[0013] Step S3, Cell Status Calculation: The battery management controller calculates the cell status information of the cell based on the cell parameters, the scene parameters, and the cell status calculation algorithm;

[0014] Step S4, Dynamic Equalization Control: The battery management controller dynamically adjusts the equalization threshold and the equalization rate based on the cell status information, the scenario parameters, and the aging stage of the battery, and executes the equalization strategy according to the adjusted equalization threshold and equalization rate to obtain equalization status information.

[0015] Step S5, Algorithm self-learning iteration: The battery management system periodically uploads the running data stored in the storage unit from the controller to the cloud server, and receives the algorithm update package sent by the cloud server to update the adaptive algorithm library;

[0016] Step S6, battery pack status determination: The battery management master controller receives the cell status information and the equalization status information reported by each battery management slave controller, and calculates the overall status information of the battery pack.

[0017] According to certain preferred embodiments of the present invention, in step S1, the scene parameters include at least temperature, humidity and air pressure.

[0018] According to certain preferred embodiments of the present invention, in step S1, evaluating the aging stage of the battery based on the historical charge-discharge data includes: dividing the aging stage of the battery into a new battery stage, a semi-aged stage, and a near-discard stage based on the historical charge-discharge cycle count, capacity decay rate, and internal resistance change trend.

[0019] According to certain preferred embodiments of the present invention, the criteria for determining the new battery stage are: historical charge-discharge cycle count < 50 times, capacity decay rate < 5%, and internal resistance change trend < 10%; the criteria for determining the semi-aging stage are: 50 times ≤ historical charge-discharge cycle count < 500 times, 5% ≤ capacity decay rate < 20%, and 10% ≤ internal resistance change trend < 30%; the criteria for determining the near-discard stage are: historical charge-discharge cycle count ≥ 500 times, or capacity decay rate ≥ 20%, or internal resistance change trend ≥ 30%.

[0020] According to certain preferred embodiments of the present invention, in step S2, the cell state calculation algorithm includes at least one or more of the following: low temperature scenario optimization algorithm, high temperature protection algorithm, aging battery adaptation algorithm, and high altitude low voltage algorithm, preferably two.

[0021] According to certain preferred embodiments of the present invention, in step S4, the dynamic adjustment of the equalization threshold includes setting different maximum equalization thresholds according to the aging stage of the battery, wherein the new battery stage is ≤20mV, the semi-aged stage is ≤30mV, and the near-scrap stage is ≤50mV.

[0022] According to certain preferred embodiments of the present invention, in step S4, the dynamic adjustment of the equalization rate includes adjusting the ratio of the equalization rate to the normal rate according to the temperature in the scene parameters, wherein: when the temperature is <-10℃, the equalization rate is 30%-50% of the normal rate; when the temperature is between -10℃ and 45℃, the equalization rate adopts the normal rate; when the temperature is >45℃, the equalization rate is 60%-80% of the normal rate.

[0023] According to certain preferred embodiments of the present invention, the dynamic adjustment of the equalization rate further includes dynamically adjusting the equalization rate according to the air pressure adjustment in the scene parameters, wherein: when the air pressure is <80kPa, the equalization rate is 70%-90% of the normal rate.

[0024] According to another aspect of the present invention, a distributed battery management system for the aforementioned optimization method is provided, the system comprising a battery management master controller and a plurality of battery management slave controllers, wherein the battery management master controller is configured to receive cell status information and equalization status information reported by each battery management slave controller to calculate the overall status information of the battery pack, and each of the battery management slave controllers comprises:

[0025] The data acquisition module is used to collect the cell parameters of the battery cells in real time.

[0026] The storage unit, connected to the data acquisition module, the aging feature extraction unit and the second communication module, is used to store the cell parameters and the cell's historical charge and discharge data.

[0027] The scene perception and aging status assessment module integrates an environmental sensor and an aging feature extraction unit. The environmental sensor is used to collect scene parameters, and the aging feature extraction unit is used to read the historical charge and discharge data of the battery cell from the storage unit and assess the aging stage of the battery.

[0028] An adaptive algorithm library, connected to the scene perception and aging state assessment module, is used to store cell state calculation algorithms;

[0029] A cell status calculation module, connected to the data acquisition module and the adaptive algorithm library, is used to calculate cell status information based on the cell parameters, the scene parameters and the cell status calculation algorithm.

[0030] The equalization control module is connected to the scene perception and aging state assessment module and the cell state calculation module. It is used to dynamically adjust the equalization threshold and the equalization rate according to the cell state information, the scene parameters and the aging stage of the battery, and execute the equalization strategy according to the adjusted equalization threshold and equalization rate to obtain equalization state information.

[0031] The first communication module is connected to the cell state calculation module and the equalization control module, and is used to send the calculated cell state information and the equalization state information generated by the execution to the battery management main controller.

[0032] The second communication module is connected to the storage unit and the cloud server, and is used to periodically upload the data in the storage unit to the cloud server, and receive the algorithm update package sent by the cloud server to update the adaptive algorithm library.

[0033] According to certain preferred embodiments of the present invention, the environmental sensor includes a temperature sensor, a humidity sensor, and a barometric pressure sensor.

[0034] According to certain preferred embodiments of the present invention, the aging feature extraction unit divides the aging stage of the battery into a new battery stage, a semi-aged stage, and a near-discard stage based on the historical charge-discharge cycle count, capacity decay rate, and internal resistance change trend. The criteria for determining the new battery stage are: historical charge-discharge cycle count < 50 times, capacity decay rate < 5%, and internal resistance change trend < 10%. The criteria for determining the semi-aged stage are: 50 cycles ≤ historical charge-discharge cycle count < 500 times, 5% ≤ capacity decay rate < 20%, and 10% ≤ internal resistance change trend < 30%. The criteria for determining the near-discard stage are: historical charge-discharge cycle count ≥ 500 times, or capacity decay rate ≥ 20%, or internal resistance change trend ≥ 30%.

[0035] According to certain preferred embodiments of the present invention, the targeted cell state calculation algorithms stored in the adaptive algorithm library include at least one or more of the following: low temperature scenario optimization algorithm, high temperature protection algorithm, aging battery adaptation algorithm, and high altitude low voltage algorithm, preferably two.

[0036] According to certain preferred embodiments of the present invention, the equalization control module dynamically adjusts the equalization threshold by setting different maximum equalization thresholds according to the battery aging stage, wherein the new battery stage is ≤20mV, the semi-aged stage is ≤30mV, and the near-discard stage is ≤50mV.

[0037] According to certain preferred embodiments of the present invention, the equalization control module dynamically adjusts the equalization rate based on the ratio of the temperature equalization rate to the normal rate in the scene parameters, wherein: when the temperature is <-10℃, the equalization rate is 30%-50% of the normal rate; when the temperature is between -10℃ and 45℃, the normal rate is used; and when the temperature is >45℃, the equalization rate is 60%-80% of the normal rate.

[0038] According to certain preferred embodiments of the present invention, the equalization control module further adjusts the equalization rate according to the air pressure in the scene parameters, wherein: when the air pressure is <80kPa, the equalization rate is 70%-90% of the normal rate.

[0039] By employing the method and system of this invention, and by adding a scene perception and aging state assessment module, an adaptive algorithm library, a second communication module, and a storage unit, functions such as dynamic algorithm invocation, dynamic equilibrium control, and algorithm self-learning iteration are realized. Compared with existing technologies, this can effectively improve the calculation accuracy of cell status, extend battery cycle life, and improve the safety of battery use. Attached Figure Description

[0040] The accompanying drawings are provided in this specification to more clearly explain the technical solutions of the present invention; however, the art is not limited thereto. The flowcharts shown in the drawings are merely illustrative and do not necessarily include all contents and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be decomposed, combined, or partially merged, so the actual execution order may change according to the actual situation.

[0041] Figure 1 This is a flowchart illustrating an optimization method for a distributed battery management system according to one embodiment of the present invention.

[0042] Figure 2 This is a schematic diagram of a distributed battery management system according to one embodiment of the present invention;

[0043] Figure 3This is a schematic diagram of the internal structure of the battery management controller in a distributed battery management system according to one embodiment of the present invention.

[0044] Explanation of reference numerals in the attached figures:

[0045] S1-S6. Flowchart of the optimization method for the distributed battery management system; 20. Main controller of the distributed battery management system; 201. Third communication module; 30. Battery management slave controller structure of the distributed battery management system; 301. Data acquisition module; 302. Storage unit; 303. Scene perception and aging state assessment module; 303a. Environmental sensor; 303b. Aging feature extraction unit; 304. Adaptive algorithm library; 305. Cell state calculation module; 306. Equalization control module; 307. First communication module; 308. Second communication module; 40. Cloud server. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0047] In this document, unless otherwise specified, the term "balancing threshold" refers to the critical value of the voltage difference between battery cells that triggers the battery management system to activate the balancing function. When the voltage difference between the highest-voltage cell and the lowest-voltage cell in the battery module exceeds this threshold, the system determines that the "inconsistency" between the cells is too large to require intervention, thus initiating balancing. If the voltage difference does not exceed this threshold, balancing is unnecessary to avoid unnecessary energy loss and cell stress.

[0048] In this document, unless otherwise specified, the term "equalization rate" refers to the magnitude of the current used to discharge (or charge) a high-voltage cell during the equalization process. The equalization rate determines the "intensity" and "speed" of the equalization process. An excessively high rate may lead to localized overheating, lithium plating, or additional capacity loss; an excessively low rate results in low equalization efficiency and may fail to effectively suppress inconsistencies.

[0049] Example 1

[0050] like Figure 1 , Figure 2 and Figure 3 As shown, the optimization method of the distributed battery management system of the present invention will be described in detail below. In this embodiment, the distributed battery management system of a pure electric passenger vehicle is used as an example, and the specific configuration is as follows:

[0051] Battery pack type: Ternary lithium battery pack, nominal total voltage 350V, total energy 70kWh (corresponding to nominal capacity of 200Ah);

[0052] Battery management main controller: 1 unit, supporting CAN FD communication;

[0053] Battery management controllers: 12 units, each managing a module consisting of 18 cells connected in series (216 cells in total), supporting CAN communication and Ethernet communication;

[0054] Cloud server: Employs high-performance computing servers to support distributed data storage and GBDT algorithm analysis;

[0055] Environmental sensors: Each battery management unit integrates one set of temperature sensors (-40℃-85℃), humidity sensors (0% RH-100% RH), and barometric pressure sensors (30kPa-110kPa) from the controller.

[0056] Normal equalization rate: set to 0.2C (i.e. 40A, calculated based on a battery pack nominal capacity of 200Ah).

[0057] Data upload cycle: 24 hours;

[0058] Algorithm update cycle: once a week.

[0059] The specific optimization scheme for the distributed battery management system is as follows:

[0060] Step S1, Data Acquisition and Status Assessment: The battery management system acquires cell parameters from the controller in real time, acquires scene parameters in real time through integrated environmental sensors, and reads historical charge and discharge data of the cells from its own storage unit through the aging feature extraction unit and assesses the aging stage of the battery based on the historical charge and discharge data.

[0061] The scene parameters collected by the environmental sensors are: temperature -18℃, humidity 45%RH, and air pressure 100kPa.

[0062] The data acquisition module collects the parameters of each cell in real time: cell voltage 3.58V-3.62V, average voltage 3.60V, current 20A (driving discharge state), internal resistance 65mΩ-68mΩ, average internal resistance 66.5mΩ;

[0063] The data acquisition module reads historical data from the storage unit: the battery pack has been used for 3 months, with 28 historical charge-discharge cycles, an initial capacity of 200Ah, a current capacity of 198Ah, a capacity decay rate of (200-198) / 200×100%=1.0%, an initial average internal resistance of 65mΩ, a current average internal resistance of 66.5mΩ, and an internal resistance change trend of (66.5-65) / 65×100%≈2.31%.

[0064] The aging feature extraction unit extracts aging features. Cycle life decay coefficient = 28 / 500 = 0.056; capacity decay rate = 1.0% / 3 months ≈ 0.33% / month; Criteria: Cycle count < 50 cycles (28 cycles < 50 cycles), capacity decay rate < 5% (1.0% < 5%), internal resistance change trend < 10% (2.31% < 10%). All conditions for the "new battery stage" are met, i.e., cycle count < 50 cycles, capacity decay rate < 5%, and internal resistance change trend < 10%.

[0065] The scene perception and aging status assessment module synchronously sends scene parameters (temperature -18℃, humidity 45% RH, air pressure 100kPa) and aging stage (new battery stage) to the adaptive algorithm library and equalization control module, and stores them in the storage unit.

[0066] Step S2, Dynamic Algorithm Invocation: The battery management controller dynamically invokes a matching cell state calculation algorithm from the adaptive algorithm library based on the scenario parameters and the aging stage of the battery.

[0067] The adaptive algorithm library receives scenario parameters (temperature -18℃ < -10℃, which is a low temperature scenario) and aging stage (new battery stage).

[0068] Through a two-dimensional matching mechanism of "scenario-aging stage", a fusion model of "low temperature scenario optimization algorithm + new battery standard algorithm" is invoked, specifically:

[0069] Low-temperature scenario optimization algorithm: Adjust the temperature compensation coefficient for SOC calculation (from the usual 0.005 / ℃ to 0.008 / ℃), and optimize the low-temperature energy loss model for SOE calculation (introduce a low-temperature discharge efficiency coefficient of 0.85).

[0070] The new standard battery algorithm uses nominal internal resistance (65mΩ) and nominal capacity (200Ah) for calculation.

[0071] The adaptive algorithm library sends the fusion algorithm model to the cell status calculation module.

[0072] Step S3, Cell Status Calculation: The battery management controller calculates the cell status information of the cell based on the cell parameters, the scene parameters, and the called cell status calculation algorithm;

[0073] The cell status calculation module receives cell parameters (voltage 3.60V, current 20A, internal resistance 66.5mΩ) and fusion algorithm model from the data acquisition module;

[0074] SOC: The current SOC is calculated to be 65% using the ampere-hour integral method combined with the low temperature compensation coefficient.

[0075] SOH: Based on a capacity decay rate of 1.0% and an internal resistance change trend of 2.31%, combined with the evaluation of the new battery standard, the calculated SOH is 99.0%.

[0076] SOP: Based on the current SOC 65%, SOH 99.0%, temperature -18℃, internal resistance 66.5mΩ, and combined with the low temperature power limitation model, the peak discharge power is calculated to be 135kW and the peak charging power is 90kW.

[0077] SOE: Based on SOC 65%, current capacity 198Ah, total voltage 350V, and low-temperature energy loss coefficient 0.85, the remaining usable energy is calculated to be 198Ah×350V×65%×0.85 / 1000≈38.2kWh.

[0078] The cell status calculation module sends the cell status information to the equalization control module and the storage unit.

[0079] Step S4, dynamic balancing control: The battery management controller dynamically adjusts the balancing threshold and balancing rate based on the cell status information, the scenario parameters, and the aging stage of the battery, and executes the balancing strategy according to the adjusted balancing threshold and balancing rate to obtain balancing status information.

[0080] The equalization control module receives scenario parameters (temperature -18℃, air pressure 100kPa) and aging stage (new battery stage).

[0081] Adjusting the equalization threshold: Based on the new battery stage, the consistency is good, so the equalization threshold is set to ≤20mV;

[0082] Equilibrium rate: When the temperature is -18℃ to -10℃, the low temperature protection strategy is activated, and the equilibrium rate is 50% of the normal rate of 0.2C, i.e., 0.1C (20A).

[0083] The equalization control module detects the voltage difference between each cell: the maximum voltage difference is 18mV (one cell is 3.62V, another is 3.60V), which does not exceed the 20mV threshold, so equalization control is not initiated; if the voltage of a cell subsequently drops to 3.58V, and the voltage difference reaches 40mV, then equalization control is initiated. Discharge equalization is performed on the 3.62V cell at a rate of 20A. After 100 seconds of continuous equalization control, the voltage difference drops to 10mV, and equalization is complete.

[0084] The equalization control module records the equalization execution effect data and equalization status information (equalization completed). It sends the calculated cell status information and the equalization status information generated by the execution to the main controller through the first communication module and stores them in the storage unit.

[0085] Step S5, algorithm self-learning iteration: the battery management controller periodically uploads the running data stored in the storage unit to the cloud server through the second communication module, and receives the algorithm update package sent by the cloud server to update the adaptive algorithm library.

[0086] The battery management slave controller communicates with the battery management master controller via an internal CAN network to transmit real-time control and status information; the master controller communicates with the cloud server via an on-board gateway; the battery management slave controller also has an Ethernet communication interface for periodically uploading data directly to the cloud server.

[0087] The second communication module reads historical data from the storage unit within a 24-hour cycle, including:

[0088] Scene parameters: Temperature range -20℃ to -15℃, humidity 40%-50%RH, air pressure 98kPa-101kPa.

[0089] Cell parameters: Voltage 3.55V-3.65V, Current 0-30A, Internal resistance 65mΩ-68mΩ.

[0090] State calculation results: SOC 40%-80%, SOH 98.8%-99.2%, SOP 120kW-140kW, SOE 23kWh-38kWh;

[0091] Equalization performance data: One equalization occurred, with a maximum voltage difference of 40mV, a minimum voltage difference of 10mV after equalization, an average equalization capacity loss of 0.25Ah, and a maximum temperature rise of 0.5℃.

[0092] The second communication module encrypts the data and uploads it to the cloud server via Ethernet communication;

[0093] The cloud server uses the GBDT (Gradient Boosting Decision Tree) algorithm to analyze the data and optimize the algorithm model.

[0094] It was found that in low-temperature scenarios ranging from -20℃ to -15℃, the SOC value calculated using the current optimized algorithm (temperature compensation coefficient 0.008 / ℃) had an average deviation of ±2.0% and a maximum deviation of ±2.5% compared to the calibrated open-circuit voltage measured after the vehicle had been stationary for 6 hours. Through GBDT model training, it was determined that optimizing the temperature compensation coefficient in the low-temperature scenario optimization algorithm from 0.008 / ℃ to 0.0082 / ℃ could reduce the average deviation to ±1.5% and the maximum deviation to ±2.0%.

[0095] The cloud server generates an algorithm update package (including an optimized temperature compensation coefficient parameter table and version number), and pushes it to the vehicle through the vehicle gateway during the weekly update cycle, and then distributes it to the second communication module of each battery management slave controller via the CAN network;

[0096] After receiving the update packet, the second communication module forwards it to the adaptive algorithm library to complete the algorithm iteration update. The new version of the algorithm will be automatically called the next time the same low temperature scenario is entered, and the SOC calculation accuracy will be significantly improved.

[0097] Step S6: Battery pack status is determined. The battery management master controller receives the cell status information and the equalization status information reported by each battery management slave controller, and calculates the overall status information of the battery pack.

[0098] The third communication module of the main controller receives cell status information (SOC, SOH, SOP, SOE) and equalization status information sent from the controller by 12 battery managers;

[0099] The battery pack status fusion calculation module uses a weighted fusion algorithm to calculate the overall pack status:

[0100] Total Pack SOC: Calculated using a weighted average method based on the current capacity of each cell. For example, the calculated total pack SOC is 64.5%.

[0101] Overall SOH of the package: A comprehensive evaluation method of "minimum value priority, average value reference" is adopted. For example, the minimum SOH of each cell (98.8%) and the average value (99.0%) are taken, and the weighted average SOH of the package is calculated to be 98.9%.

[0102] SOP (Peak Discharge Power) for the entire battery pack: Following the "weakest link" principle, the minimum peak discharge power calculated in real-time for all cells is taken, for example, 120kW. The same applies to peak charging power; the minimum value is taken, for example, 85kW.

[0103] Total Pack SOE: The remaining usable energy of the entire pack is obtained by summing the module SOE (usable energy corrected based on the low-temperature efficiency coefficient) reported by each battery management from the controller. For example, 37.8kWh.

[0104] Overall package consistency assessment: Calculate the standard deviation (e.g., 5mV) and maximum value difference (e.g., 15mV) of the voltage of all cells, and judge the overall consistency to be good;

[0105] The main controller sends the above-mentioned overall package status information to the vehicle controller via CAN FD. Based on this, the vehicle controller formulates the vehicle energy management strategy, as follows:

[0106] Normal mode: Current SOC=64.5%, SOH=98.9%, maximum allowable discharge power 120kW, vehicle supports full-function driving;

[0107] Low battery protection mode: When the total SOC of the battery pack is below 20%, the vehicle controller will limit the maximum discharge power to 60kW and illuminate the charging indicator light on the instrument panel;

[0108] Abnormal Handling Mode: If the main controller detects that the SOH of a cell reported by the controller for a certain battery management unit drops rapidly below 97.0%, or that the internal resistance changes by more than 15%, it will immediately send an alarm signal to the vehicle controller. The vehicle controller will forcibly limit the charging and discharging power to 30kW and display a warning message "Battery system detection in progress, please drive safely" on the instrument panel. At the same time, it will upload detailed fault codes to the cloud and prompt the background to arrange maintenance.

[0109] By adopting the optimization method of the present invention, each battery management controller can autonomously call the matching algorithm and dynamically adjust the balancing strategy according to real-time environmental parameters and cell aging stage, which significantly improves the adaptability and reliability of the system under complex working conditions.

[0110] Example 2

[0111] like Figure 1 , Figure 2 and Figure 3 As shown, this embodiment is a distributed battery management system for an energy storage system deployed in a high-altitude area, with the following specific configuration:

[0112] Battery pack type: Lithium iron phosphate battery pack, nominal total voltage 1000V, total energy 500kWh (corresponding to nominal capacity of 500Ah).

[0113] Battery management main controller: 1 unit, supporting Ethernet communication;

[0114] Battery management slave controllers: 20 units, each managing a module consisting of 24 battery cells connected in series (480 cells in total). The battery management slave controllers communicate with the main controller via an internal CAN network and are equipped with a first communication module for reporting status information; simultaneously, each battery management slave controller has a built-in second communication module supporting 5G communication for direct data interaction with the cloud server.

[0115] Cloud server: Employs high-performance computing servers to support neural network algorithm (such as LSTM) analysis;

[0116] Environmental sensors: Each battery management unit integrates one set of temperature sensors (-20℃-85℃), humidity sensors (0% RH-100% RH), and barometric pressure sensors (30kPa-110kPa) from the controller.

[0117] Normal equalization rate: set to 0.15C (i.e. 75A, calculated based on the nominal capacity of the battery pack of 500Ah).

[0118] Data upload cycle: 12 hours;

[0119] Algorithm update cycle: once every 3 days.

[0120] The specific scheme for the optimization method using a distributed battery management system is as follows:

[0121] Step S1, Data Acquisition and Status Assessment: The battery management system collects cell parameters in real time from the controller, collects scene parameters in real time through integrated environmental sensors, reads historical charge and discharge data of the cells from its own storage unit through the aging feature extraction unit, and assesses the aging stage of the battery based on the historical charge and discharge data.

[0122] The energy storage system is deployed in a high-altitude area in western China (4,000 meters above sea level). The environmental parameters collected by the environmental sensors are: temperature 28°C, humidity 60% RH, and air pressure 62 kPa.

[0123] The data acquisition module collects the parameters of each cell in real time: cell voltage 3.20V-3.25V, average voltage 3.22V, current 100A (discharge state), internal resistance 88mΩ-95mΩ, average internal resistance 92mΩ;

[0124] The data acquisition module reads historical data from the storage unit: The energy storage system has been used for 18 months, with 420 historical charge-discharge cycles, an initial capacity of 500Ah, a current capacity of 425Ah, a capacity decay rate of (500-425) / 500×100%=15%, an initial average internal resistance of 70mΩ, a current average internal resistance of 92mΩ, and an internal resistance change trend of (92-70) / 70×100%≈31.43%;

[0125] The aging feature extraction unit extracts aging features. According to the preset rules (number of cycles ≥ 500 or capacity decay rate ≥ 20% or internal resistance change trend ≥ 30%), since the internal resistance change trend (31.43%) ≥ 30%, it is determined to be "near scrap stage".

[0126] The scene perception and aging status assessment module synchronously sends scene parameters (temperature 28℃, humidity 60% RH, air pressure 62kPa) and aging stage (near scrap stage) to the adaptive algorithm library and the equalization control module, and stores them in the storage unit.

[0127] Step S2, algorithm dynamic invocation: The battery management controller dynamically invokes a matching cell state calculation algorithm from the adaptive algorithm library based on the scenario parameters and the battery aging stage.

[0128] The adaptive algorithm library receives scene parameters (air pressure 62kPa<80kPa, belonging to high altitude low pressure scene; temperature 28℃ within the range of -10℃-45℃) and aging stage (near scrap stage);

[0129] Through a dual-dimensional matching mechanism of "scenario-aging stage", a fusion model of "high altitude low pressure algorithm + aging battery adaptation algorithm" is invoked.

[0130] High-altitude low-pressure algorithm: Adjust the air pressure influence factor in SOE calculation (from the usual 1.0 to 0.92) to compensate for the impact of heat dissipation changes on available energy under low air pressure; optimize the lithium-ion diffusion kinetics compensation coefficient in SOC calculation;

[0131] Aging battery adaptation algorithm: Adjust the internal resistance decay coefficient calculated by SOP (from the usual 1.0 to 1.3) to reflect the impact of increased internal resistance; optimize the SOH evaluation model and increase the weight of cycle number decay.

[0132] The adaptive algorithm library sends the fusion algorithm model to the cell status calculation module.

[0133] Step S3, cell status calculation: The battery management controller calculates the cell status information of the cell based on the cell parameters, the scene parameters, and the called cell status calculation algorithm.

[0134] The cell status calculation module receives cell parameters (voltage 3.22V, current 100A, internal resistance 92mΩ) and fusion algorithm model from the data acquisition module;

[0135] SOC calculation: Using the ampere-hour integral method combined with the pressure correction factor, the current SOC is calculated to be 75%.

[0136] SOH calculation: Based on a capacity decay rate of 15%, an internal resistance change trend of 31.43%, and 420 cycles, combined with a comprehensive evaluation model of aging battery adaptation algorithm, the SOH was calculated to be 84.8%.

[0137] SOP calculation: Based on the current SOC 75%, SOH 84.8%, temperature 28℃, internal resistance 92mΩ and internal resistance decay coefficient 1.3, the peak discharge power is calculated to be 350kW and the peak charging power is 280kW.

[0138] SOE Calculation: Based on SOC 75%, current capacity 425kWh, and pressure influence factor 0.92, the remaining usable energy is calculated to be 425Ah×1000V×75%×0.92 / 1000≈293.25 kWh;

[0139] The cell status calculation module sends the cell status information to the equalization control module and the storage unit.

[0140] Step S4, dynamic balancing control: The battery management controller dynamically adjusts the balancing threshold and balancing rate based on the cell status information, the scenario parameters, and the aging stage of the battery, and executes the balancing strategy according to the adjusted threshold and rate to obtain balancing status information.

[0141] The equalization control module receives scene parameters (temperature 28℃, air pressure 62kPa) and aging stage (near scrap stage).

[0142] Adjust the equalization threshold: Based on the near-scrap stage, set the equalization threshold to ≤50mV;

[0143] Adjust the equalization rate: The air pressure is 62 kPa < 80 kPa, so the equalization rate is 80% of the normal rate of 0.15C, that is, 0.12C (60A).

[0144] The equalization control module detects the voltage difference between each cell: the maximum voltage difference is 48mV (one cell is 3.25V, another cell is 3.17V), reaching the equalization threshold of 50mV, and the equalization strategy is activated, that is, the cell with voltage of 3.25V is discharged for equalization, and the cell with voltage of 3.17V is charged for equalization, with an equalization rate of 60A; after equalization continues for 180 seconds, the cell voltage difference drops to 20mV, and equalization is completed.

[0145] The equalization control module records the equalization execution effect data and equalization status information (equalization completed). It sends the cell status information and equalization status information to the main controller through the first communication module and stores them in the storage unit.

[0146] Step S5, algorithm self-learning iteration: the battery management controller periodically uploads the running data stored in the storage unit to the cloud server through the second communication module, and receives the algorithm update package sent by the cloud server to update the adaptive algorithm library.

[0147] The second communication module reads historical data from the storage unit in a 12-hour cycle, including: scene parameters, cell parameters, status calculation results, and equalization execution effect data; after encrypting the data, the second communication module uploads it to the cloud server via 5G communication.

[0148] The cloud server uses an LSTM neural network algorithm to analyze the data: it was found that the deviation between the calculated SOE value and the actual discharge energy in high-altitude, low-pressure scenarios is about ±2.5%. By optimizing the air pressure influence factor of the high-altitude, low-pressure algorithm (adjusted from 0.92 to 0.91), the deviation is expected to be reduced to ±2.10%.

[0149] The cloud server generates an algorithm update package (containing optimized air pressure influence factor and internal resistance attenuation coefficient) and pushes it to the second communication module within the update cycle;

[0150] After receiving the update packet, the second communication module forwards it to the adaptive algorithm library to complete the algorithm iteration update. The calculation accuracy of SOE and SOP is significantly improved the next time the same high-altitude scene is entered.

[0151] Step S6: Battery pack status is determined. The battery management master controller receives the cell status information and the equalization status information reported by each battery management slave controller, and calculates the overall status information of the battery pack.

[0152] The third communication module of the main controller receives cell status information and equalization status information sent from the controller by 20 battery management units;

[0153] The battery pack status fusion calculation module uses a weighted fusion algorithm to calculate the overall pack status:

[0154] Overall Pack SOC: Based on the weighted average of the SOC of each cell managed by the battery management controller, the overall pack SOC is calculated to be 74.7%.

[0155] Total SOH of the entire package: Based on the minimum SOH of 84.5% and the average SOH of 84.8% for each cell, the total SOH of the entire package is calculated to be 84.6%.

[0156] Total SOP: Based on the minimum SOP of each cell, which is 300kW, the peak discharge power of the entire pack is calculated to be 300kW.

[0157] Total SOE of the entire battery pack: Based on the sum of the SOE of each cell, the remaining usable energy of the entire pack is calculated to be 293.8 kWh;

[0158] The main controller sends the overall status information to the energy storage system controller, which adjusts the charging and discharging strategy based on this information: the current SOC is 74.7%, and the maximum allowable discharge power is 300kW to meet the load demand during peak industrial and commercial electricity consumption; when the SOC is below 30%, grid charging is started with a charging power of 280kW.

[0159] If a cell managed by a battery management controller malfunctions (e.g., the cell voltage suddenly drops to 3.0V, or the state of harmonics (SOH) drops below 80%), the main controller will trigger an alarm signal, the energy storage system controller will stop the charging and discharging of the module containing that cell, and notify maintenance personnel for repair.

[0160] By adopting this solution, the SOE calculation accuracy in high-altitude and low-pressure scenarios can be effectively improved, the battery cycle life can be extended, the energy utilization efficiency of the energy storage system can be effectively improved, and the energy storage benefits can be significantly enhanced.

[0161] Example 3

[0162] like Figure 2 and Figure 3 As shown below, a distributed battery management system employing the above-described optimization method is described in detail. The system includes a battery management master controller and multiple battery management slave controllers. Each battery management slave controller includes the following functional modules, which work together to execute the above-described optimization method.

[0163] The data acquisition module is used to collect the cell parameters of the battery cells in real time.

[0164] A storage unit, connected to the data acquisition module, the aging feature extraction unit and the second communication module, is used to store the cell parameters and the cell's historical charge and discharge data.

[0165] The scene perception and aging status assessment module integrates environmental sensors and an aging feature extraction unit. The environmental sensors collect scene parameters, and the aging feature extraction unit reads historical charge / discharge data of the battery cell from the storage unit and assesses the battery's aging stage. The environmental sensors include a temperature sensor, a humidity sensor, and a barometric pressure sensor. The aging feature extraction unit divides the battery's aging stage into new battery stage, semi-aged stage, and near-retirement stage based on historical charge / discharge cycle count, capacity decay rate, and internal resistance change trends.

[0166] The criteria for determining the new battery stage are: historical charge-discharge cycle count < 50 times, capacity decay rate < 5%, and internal resistance change trend < 10%; the criteria for determining the semi-aged stage are: 50 cycles ≤ historical charge-discharge cycle count < 500 times, 5% ≤ capacity decay rate < 20%, and 10% ≤ internal resistance change trend < 30%; the criteria for determining the near-discard stage are: historical charge-discharge cycle count ≥ 500 times, or capacity decay rate ≥ 20%, or internal resistance change trend ≥ 30%.

[0167] An adaptive algorithm library, connected to the scene perception and aging state assessment module, is used to store cell state calculation algorithms; wherein, the targeted cell state calculation algorithms stored in the adaptive algorithm library include at least two of the following: low temperature scene optimization algorithm, high temperature protection algorithm, aging battery adaptation algorithm, and high altitude low voltage algorithm.

[0168] The cell status calculation module is connected to the data acquisition module and the adaptive algorithm library, and is used to calculate cell status information such as SOC, SOH, SOP, and SOE based on the cell parameters, the scenario parameters and the cell status calculation algorithm matched with the aging stage.

[0169] The equalization control module, connected to the scene perception and aging state assessment module and the cell state calculation module, is used to dynamically adjust the equalization threshold and the equalization rate based on the cell state information, the scene parameters, and the aging stage of the battery, and execute the equalization strategy according to the adjusted threshold and rate to obtain equalization state information; wherein,

[0170] The equalization control module dynamically adjusts the equalization threshold by setting different maximum equalization thresholds according to the battery aging stage, for example, ≤20mV for the new battery stage, ≤30mV for the semi-aged stage, and ≤50mV for the near-scrap stage.

[0171] The equalization control module dynamically adjusts the equalization rate based on the temperature in the scene parameters, adjusting the equalization rate relative to the normal rate. When the temperature is <-10℃, the equalization rate is 30%-50% of the normal rate; when the temperature is between -10℃ and 45℃, the normal rate is used; and when the temperature is >45℃, the equalization rate is 60%-80% of the normal rate. The equalization control module also adjusts the equalization rate based on the air pressure in the scene parameters. When the air pressure is <80kPa, the equalization rate is 70%-90% of the normal rate.

[0172] The first communication module is connected to the cell state calculation module and the equalization control module, and is used to send the calculated cell state information and the equalization state information generated by the execution to the battery management main controller.

[0173] The second communication module is connected to the storage unit and the cloud server, and is used to periodically upload the data in the storage unit to the cloud server, and receive the algorithm update package sent by the cloud server to update the adaptive algorithm library.

[0174] During system operation, the data acquisition module and environmental sensors continuously acquire data; the scene perception and aging state assessment module evaluates the current state and outputs key decision parameters (scene and aging stage); the adaptive algorithm library provides accurate algorithm models accordingly; the cell state calculation module and the equalization control module use this model to perform core calculations and control; the first and second communication modules respectively complete the internal system state reporting and external data interaction and learning. Through the above modular design and collaborative process, this system realizes the hardware and software support for the optimization method of the distributed battery management system.

[0175] This system introduces a scene awareness and aging state assessment module and an adaptive algorithm library, enabling each battery management controller to autonomously call the optimal algorithm based on the real-time environment and cell health status, thus solving the problem of poor adaptability of fixed algorithms. Through dynamic strategy adjustment by the balancing control module, it achieves refined balancing management under different aging stages and extreme scenarios, effectively maintaining module consistency and improving safety. A closed loop formed by the second communication module and the cloud server gives the system self-learning and continuous performance optimization capabilities throughout its entire lifecycle. Therefore, this system significantly improves the accuracy, reliability, and battery life of the distributed BMS under complex operating conditions.

[0176] It should be noted that the logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer system, a system including a processor or other system that can fetch and execute instructions from an instruction execution system, apparatus or device).

[0177] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0178] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0179] It should also be further understood that the term "and / or" as used in this specification refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes such combinations.

[0180] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. An optimization method for a distributed battery management system, characterized in that, The method includes the following steps: Step S1, Data Acquisition and Status Assessment: The battery management system collects cell parameters from the controller in real time, and collects scene parameters of the cell in real time through environmental sensors. The scene parameters include at least temperature, humidity, and air pressure. The system also reads historical charge and discharge data of the cell from the storage unit through the aging feature extraction unit and assesses the aging stage of the battery based on the historical charge and discharge data. Specifically, based on the historical charge and discharge cycle count, capacity decay rate, and internal resistance change trend, the aging stage of the battery is divided into new battery stage, semi-aged stage, and near-discard stage. Step S2, Dynamic Algorithm Invocation: The battery management controller dynamically invokes a matching cell state calculation algorithm from the adaptive algorithm library based on the scenario parameters and the aging stage of the battery; wherein, the cell state calculation algorithm includes at least one or more of the following: low temperature scenario optimization algorithm, high temperature protection algorithm, aging battery adaptation algorithm, and high altitude low pressure algorithm; Step S3, Cell Status Calculation: The battery management controller calculates the cell status information of the cell based on the cell parameters, the scene parameters, and the cell status calculation algorithm; Step S4, Dynamic Equalization Control: The battery management controller dynamically adjusts the equalization threshold and the equalization rate based on the cell status information, the scenario parameters, and the aging stage of the battery, and executes the equalization strategy according to the adjusted equalization threshold and equalization rate to obtain equalization status information. Step S5, Algorithm self-learning iteration: The battery management system periodically uploads the running data stored in the storage unit from the controller to the cloud server, and receives the algorithm update package sent by the cloud server to update the adaptive algorithm library; Step S6, Battery Pack Status Determination: The battery management master controller receives the cell status information and the equalization status information reported by each battery management slave controller, and calculates the overall status information of the battery pack.

2. The method according to claim 1, characterized in that, In step S4, the dynamic adjustment of the equalization threshold includes setting different maximum equalization thresholds according to the aging stage of the battery, wherein the new battery stage is ≤20mV, the semi-aged stage is ≤30mV, and the near-scrap stage is ≤50mV.

3. The method according to claim 1, characterized in that, In step S4, the dynamic adjustment of the equalization rate includes adjusting the ratio of the equalization rate to the normal rate according to the temperature in the scene parameters, wherein: when the temperature is <-10℃, the equalization rate is 30%-50% of the normal rate; when the temperature is between -10℃ and 45℃, the equalization rate is the normal rate; when the temperature is >45℃, the equalization rate is 60%-80% of the normal rate.

4. A distributed battery management system for use in the optimization method according to any one of claims 1-3, characterized in that, The system includes a battery management master controller and multiple battery management slave controllers. The battery management master controller receives cell status information and equalization status information reported by each battery management slave controller to calculate the overall status information of the battery pack. Each battery management slave controller includes: The data acquisition module is used to collect the cell parameters of the battery cells in real time. The storage unit, connected to the data acquisition module, the aging feature extraction unit and the second communication module, is used to store the cell parameters and the cell's historical charge and discharge data. The scene perception and aging status assessment module integrates an environmental sensor and an aging feature extraction unit. The environmental sensor is used to collect scene parameters, which include at least temperature, humidity, and air pressure. The aging feature extraction unit is used to read the historical charge and discharge data of the battery cell from the storage unit and assess the aging stage of the battery. The aging feature extraction unit divides the aging stage of the battery into a new battery stage, a semi-aged stage, and a near-discard stage based on the historical charge and discharge cycle number, capacity decay rate, and internal resistance change trend. An adaptive algorithm library, connected to the scene perception and aging state assessment module, is used to store cell state calculation algorithms; wherein, the cell state calculation algorithm includes at least one or more of the following: low temperature scene optimization algorithm, high temperature protection algorithm, aging battery adaptation algorithm, and high altitude low voltage algorithm; A cell status calculation module, connected to the data acquisition module and the adaptive algorithm library, is used to calculate cell status information based on the cell parameters, the scene parameters and the cell status calculation algorithm. The equalization control module is connected to the scene perception and aging state assessment module and the cell state calculation module. It is used to dynamically adjust the equalization threshold and the equalization rate according to the cell state information, the scene parameters and the aging stage of the battery, and execute the equalization strategy according to the adjusted equalization threshold and equalization rate to obtain equalization state information. The first communication module is connected to the cell state calculation module and the equalization control module, and is used to send the calculated cell state information and the equalization state information generated by the execution to the battery management main controller. The second communication module is connected to the storage unit and the cloud server, and is used to periodically upload the data in the storage unit to the cloud server, and receive the algorithm update package sent by the cloud server to update the adaptive algorithm library.

5. The system according to claim 4, characterized in that, The equalization control module dynamically adjusts the equalization threshold according to different maximum equalization thresholds set for the battery aging stage, wherein the new battery stage is ≤20mV, the semi-aged stage is ≤30mV, and the near-discard stage is ≤50mV.

6. The system according to claim 4, characterized in that, The equalization control module dynamically adjusts the equalization rate based on the ratio of the temperature adjustment rate to the normal rate in the scenario parameters. Specifically: when the temperature is less than -10°C, the equalization rate is 30%-50% of the normal rate; when the temperature is between -10°C and 45°C, the normal rate is used; and when the temperature is greater than 45°C, the equalization rate is 60%-80% of the normal rate.

7. The system according to claim 6, characterized in that, The equalization control module also adjusts the equalization rate based on the air pressure in the scene parameters.

Images