Multi-type tiered battery management method, system, apparatus, and storage medium

Abstract

The application provides a multi-type cascade battery management method, system, device and storage medium. The method comprises: sequentially polling, by an energy management system, a communication protocol of a battery cluster connected with each energy storage converter, establishing an independent communication link with each battery cluster; periodically collecting, based on the independent communication link, running state data of the battery cluster uploaded by each energy storage converter, and performing state of charge calculation to obtain a real-time state of charge value; when the statistical dispersion exceeds a preset dispersion threshold, calculating individual deviation amounts of each state of charge value and a preset equalization reference, integrating all the individual deviation amounts to obtain a personalized power bias amount set; combining all the real-time state of charge values and the personalized power bias amount set to perform power optimization distribution, generating and issuing independent control instructions to each energy storage converter to complete SOC equalization management. The application can flexibly adjust the power distribution strategy according to the real-time state.

Description

Technical Field

[0001] This application relates to the field of battery management technology, and in particular to methods, systems, devices and storage media for managing various types of cascaded batteries. Background Technology

[0002] With the increasing proportion of renewable energy and the large-scale retirement of electric vehicles, the cascade utilization of power batteries to build industrial and commercial energy storage systems has become an important technical path for peak shaving and valley filling and improving grid stability. Against this backdrop, how to efficiently and safely manage multiple retired battery clusters with different degradation characteristics and models, especially achieving precise state-of-charge balancing, has become a key constraint on the large-scale application of this technology. Existing collaborative management schemes simply connect multiple battery clusters in parallel and centrally control them with a single large converter. This approach ignores the significant differences between individual batteries in the cascade system and makes it difficult to perform fine-grained energy scheduling based on the health status of each battery cluster. This extensive management method may lead to a decrease in the overall available capacity of the system, accelerated degradation of individual battery cells, and even safety hazards, thereby affecting the overall economic efficiency and reliability of the energy storage system. Summary of the Invention

[0003] To overcome the problems existing in related technologies, this application provides a multi-type cascade battery management method, system, device and storage medium, which can flexibly adjust the power allocation strategy according to the real-time status, making the system more adaptable to the collaborative management scenario of multi-type and multi-degradation cascade batteries.

[0004] This application provides a multi-type cascade battery management method, including: The energy management system sequentially polls the battery clusters connected to each energy storage converter using the communication protocol to establish an independent communication link with each battery cluster. Based on the independent communication link, the battery cluster operation status data uploaded by each of the energy storage converters is periodically collected, and the state of charge is calculated to obtain the real-time state of charge value. Calculate the statistical dispersion of the state of charge values. When the statistical dispersion exceeds a preset dispersion threshold, calculate the individual deviation of each state of charge value from the preset equilibrium benchmark. Integrate all the individual deviations to obtain a personalized power bias set. In response to the power dispatching requirements of the mains power, power optimization allocation is performed by combining all the real-time state of charge values ​​and the personalized power bias set, generating and issuing independent control commands to each of the energy storage converters to complete SOC equalization management; The step of polling the battery clusters connected to each energy storage converter sequentially using the communication protocol through the energy management system to establish an independent communication link with each battery cluster includes: For a given battery cluster, a protocol type is selected sequentially from a pre-set communication protocol library as the current trial protocol, assembled into a corresponding handshake instruction frame, and sent to the energy storage converter connected to the battery cluster. When the response data from the energy storage converter is received within the preset timeout period, the response data is verified according to the specifications of the current trial protocol. If the response data passes verification, the frame structure of the response data is parsed to obtain the battery cluster status parameters, and the battery cluster status parameters are compared one by one with the preset status value range. When all the state parameters of the battery cluster are within the range of the state values, it is determined that the communication protocol of the battery cluster is successfully matched, and the corresponding energy storage converter is bound to the current trial protocol to complete the establishment of the independent communication link.

[0005] Furthermore, the step of periodically collecting the battery cluster operating status data uploaded by each of the energy storage converters based on the independent communication link, and performing state of charge calculation to obtain a real-time state of charge value includes: The battery cluster operating status data, including current sampling sequences and timestamp sequences, is obtained from each of the energy storage converters via the independent communication links. By traversing the current sampling sequence and the timestamp sequence, the current value of each sampling point in the current sampling sequence is multiplied by the sampling time interval in the timestamp sequence to obtain the charge change in each sampling period; The cumulative charge change rate is obtained by summing the charge changes over all the sampling periods and combining them with the rated capacity parameter of the battery cluster. The cumulative rate of change of charge is algebraically superimposed with the initial state of charge value of the battery, and the superposition result is set as the real-time state of charge value.

[0006] Further, the step of calculating the statistical dispersion of the state of charge (SCC) values, when the statistical dispersion exceeds a preset dispersion threshold, involves calculating the individual deviation of each SCC value from a preset equilibrium benchmark, and integrating all deviations to obtain a personalized power bias set, including: The real-time state of charge values ​​of all the battery clusters are collected, and the difference between the highest and lowest real-time state of charge values ​​is extracted and calculated. The difference is used as the statistical dispersion. The statistical dispersion is compared with the preset dispersion threshold. If the statistical dispersion is greater than the preset dispersion threshold, the algebraic difference between each state of charge value and the preset equilibrium benchmark is calculated, and the algebraic difference is set as the individual deviation. Calculate the arithmetic mean of all the individual deviations as a common offset, subtract the common offset from each individual deviation, and integrate the resulting individual power offsets to obtain the personalized power offset set.

[0007] Further, the step of aggregating the real-time state of charge (SOC) values ​​of all the battery clusters, extracting and calculating the difference between the highest and lowest SOC values, and using the difference as the statistical dispersion, includes: The real-time state of charge values ​​of all the battery clusters are traversed, and the real-time state of charge values ​​are sorted in descending order to form an ordered state of charge sequence. Extract the first element from the ordered sequence of charged states as the highest charged state value, and extract the last element as the lowest charged state value. The absolute difference between the highest state of charge value and the lowest state of charge value is obtained by algebraically subtracting the two values. Verify whether the number of data points in the ordered charged state sequence meets the preset minimum statistical sample requirement. When the requirement is met, output the absolute difference as the statistical dispersion.

[0008] Furthermore, in response to the power dispatching requirements of the mains power, power optimization allocation is performed by combining all the real-time state of charge values ​​and the personalized power bias set, generating and issuing independent control commands to each of the energy storage converters to complete SOC equalization management, including: Based on the real-time state of charge value, the rated power parameters of the corresponding battery cluster are obtained, and the remaining usable capacity of the battery cluster is calculated. Calculate the power ratio of the remaining available capacity to the rated power parameter, and perform an initial allocation of the power scheduling demand based on the power ratio to obtain the base power value; Extract the power adjustment value from the personalized power bias set and adjust the base power value to obtain the desired power value; Based on the rated power parameters, the desired power value is limited to obtain a set of target output power values; Based on the target output power value set, a corresponding power control command is generated for each energy storage converter, and then sent to each energy storage converter for execution through the independent communication link to complete SOC equalization management.

[0009] Further, the step of limiting the desired power value based on the rated power parameter to obtain the target output power value includes: Extract the rated power parameter corresponding to each battery cluster and compare it with the expected power value respectively; When the desired power value exceeds the upper limit of the corresponding rated power parameter, the upper limit of the rated power parameter is set as the target output power value; When the desired power value is lower than the corresponding lower limit of the rated power parameter, the lower limit of the rated power parameter is set as the target output power value; When the desired power value is within the upper and lower limits of the corresponding rated power parameter, the desired power value is directly set as the target output power value; The target output power values ​​of all the battery clusters after the limiting process are summarized to form the target output power value set.

[0010] This application also provides a multi-type cascade battery management system, applied to any of the multi-type cascade battery management methods described above, including: The acquisition module is used to poll the battery clusters connected to each energy storage converter in turn through the energy management system to establish an independent communication link with each battery cluster. The analysis module is used to periodically collect the battery cluster operating status data uploaded by each of the energy storage converters based on the independent communication link, and to perform state of charge calculation to obtain real-time state of charge values. The association module is used to calculate the statistical dispersion of the state of charge value. When the statistical dispersion exceeds a preset dispersion threshold, the module calculates the individual deviation of each state of charge value from the preset equilibrium benchmark and integrates all the individual deviations to obtain a personalized power bias set. The processing module is used to respond to the power dispatching requirements of the mains power, combine all the real-time state of charge values ​​and the personalized power bias set to perform power optimization allocation, generate and send independent control commands to each of the energy storage converters, and complete SOC balance management. The step of polling the battery clusters connected to each energy storage converter sequentially using the communication protocol through the energy management system to establish an independent communication link with each battery cluster includes: For a given battery cluster, a protocol type is selected sequentially from a pre-set communication protocol library as the current trial protocol, assembled into a corresponding handshake instruction frame, and sent to the energy storage converter connected to the battery cluster. When the response data from the energy storage converter is received within the preset timeout period, the response data is verified according to the specifications of the current trial protocol. If the response data passes verification, the frame structure of the response data is parsed to obtain the battery cluster status parameters, and the battery cluster status parameters are compared one by one with the preset status value range. When all the state parameters of the battery cluster are within the range of the state values, it is determined that the communication protocol of the battery cluster is successfully matched, and the corresponding energy storage converter is bound to the current trial protocol to complete the establishment of the independent communication link.

[0011] This application also provides a multi-type cascade battery management device, including: Memory, used to store programs; A processor is used to execute the program to implement the various steps of the multi-type cascade battery management method described in any of the above-mentioned embodiments.

[0012] This application also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0013] The technical solution provided in this application may include the following beneficial effects: This application enables refined monitoring of heterogeneous cascaded batteries by establishing independent communication links for each battery cluster and implementing "one cluster, one management," thus laying the foundation for subsequent personalized energy dispatch and effectively solving the problem of coarse management caused by inconsistent battery degradation characteristics. By calculating the statistical dispersion of the state of charge (SOC) value and using it as the balancing trigger condition, combined with individual deviations to calculate the power bias, dynamic quantitative assessment and proactive intervention of consistency among battery clusters are achieved, thereby improving the overall available capacity and balancing efficiency of the system. By responding to grid dispatch requirements and integrating real-time status and power bias sets for optimized power allocation, the internal balancing strategy can be executed while meeting external power demands, thereby improving the economy and response accuracy of the energy storage system when participating in peak-valley arbitrage, frequency regulation, and other applications. By issuing personalized power control commands to each independent energy storage converter, independent closed-loop control of the charging and discharging process of each battery cluster is achieved, effectively avoiding safety risks caused by overcharging or over-discharging of individual battery clusters and enhancing the overall reliability of the system.

[0014] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0015] The above and other objects, features and advantages of this application will become more apparent from the more detailed description of exemplary embodiments thereof in conjunction with the accompanying drawings, wherein the same reference numerals denote the same components in the exemplary embodiments thereof.

[0016] Figure 1 This application provides a flowchart of a multi-type cascade battery management method; Figure 2A structural diagram illustrating the connection between the management system, energy storage converter, and battery provided in an embodiment of this application. Figure 3 This application provides a structural diagram of a multi-type cascade battery management system; Figure 4 This application provides a structural diagram of a multi-type cascade battery management device. Detailed Implementation

[0017] Preferred embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While preferred embodiments of the present application are shown in the drawings, it should be understood that the present application may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to make the present application more thorough and complete, and to fully convey the scope of the present application to those skilled in the art.

[0018] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0019] It should be understood that although the terms "first," "second," "third," etc., may be used in this application to describe various information, this information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0020] Reference Figure 1 and Figure 2 As shown, this application provides a multi-type cascade battery management method, including: Step S10: The energy management system sequentially polls the battery clusters connected to each energy storage converter using the communication protocol to establish an independent communication link with each battery cluster. Step S20: Based on the independent communication link, periodically collect the battery cluster operation status data uploaded by each energy storage converter, and perform state of charge calculation to obtain the real-time state of charge value. Step S30: Calculate the statistical dispersion of the state of charge values. When the statistical dispersion exceeds the preset dispersion threshold, calculate the individual deviation of each state of charge value from the preset equilibrium benchmark, and integrate all individual deviations to obtain a personalized power bias set. Step S40: Respond to the power dispatching requirements of the mains power, combine all real-time state of charge values ​​and personalized power bias sets to optimize power allocation, generate and issue independent control commands to each energy storage converter, and complete SOC equalization management.

[0021] Based on the steps described above, the detailed steps are as follows: Step S10: The energy management system establishes connections with each energy storage converter (PCS1, PCS2, PCS3) via communication lines, polling the communication protocols to adapt to the communication requirements of heterogeneous battery clusters. The energy management system has a built-in pre-built communication protocol library containing various standard protocols such as CAN bus and Modbus to cover the communication specifications of different battery management systems.

[0022] For each battery cluster, the energy management system sequentially selects a protocol type from the protocol library as the current trial protocol and assembles the corresponding handshake command frame. The handshake command frame contains a command code and a verification field, and is sent via a communication line to the energy storage converter connected to the target battery cluster. Upon receiving the command, the energy storage converter forwards it to the battery management system of the battery cluster.

[0023] After sending a command, the energy management system starts a timer to wait for a response. If no response is received within the preset timeout period, the system automatically switches to the next protocol type and repeats the handshake process. When a response is received, the energy management system performs frame structure parsing and checksum verification according to the specifications of the currently used protocol to ensure data integrity.

[0024] After successful verification, the data frame is parsed to extract battery cluster status parameters, such as voltage, current, and temperature, and these parameters are compared with preset safety ranges. If all parameters are within a reasonable range, the energy management system determines that the protocol match is successful, binds the current protocol to the energy storage converter, records the communication configuration parameters, and completes the establishment of an independent communication link. In terms of connectivity, the energy management system is directly connected to each energy storage converter (PCS1, PCS2, PCS3) via communication lines, and the energy storage converters are connected to their corresponding battery clusters (Battery 1, Battery 2, Battery 3) via DC power lines, forming independent physical channels and providing a foundation for refined management.

[0025] Step S20: Based on established independent communication links, the energy management system periodically acquires battery cluster operating status data from each energy storage converter via communication lines. The acquisition period is set to a fixed interval, such as once per second, to ensure data real-time performance. The energy storage converter acquires raw data from the connected battery cluster battery management system, including current sampling sequences and timestamp sequences. The current sampling sequence contains instantaneous current values ​​at multiple sampling points, and the timestamp sequence records the time interval between each sampling point.

[0026] After receiving the data, the energy management system first verifies data integrity, such as checking sequence length and filtering outliers. It then iterates through the current sampling sequence and timestamp sequence, multiplying the current value at each sampling point by the corresponding time interval to calculate the charge change within each sampling period. The charge changes across all sampling periods are summed to obtain the total charge change. Finally, the total charge change is divided by the rated capacity parameter of the battery cluster to obtain the cumulative charge change rate.

[0027] The cumulative rate of change of charge is algebraically superimposed with the initial state of charge (SOC) value of the battery cluster, which is obtained from the battery management system (BMS) or set based on historical data during system initialization. The superposition result is the real-time SOC value. If the result exceeds a reasonable range (e.g., 0% to 100%), it is constrained to boundary values. This process achieves high-precision SOC estimation using the ampere-hour integration method, compensating for potential errors in the BMS report.

[0028] In terms of connectivity, data is transmitted from the energy storage converter to the energy management system via a communication line. The energy storage converter monitors the battery clusters in real time via a DC power line, ensuring the synchronization and accuracy of data acquisition. This step provides the core input for subsequent consistency analysis and power allocation.

[0029] Step S30: After acquiring the real-time state of charge (SOC) values ​​uploaded by each energy storage converter via communication lines, the energy management system performs a consistency assessment between battery clusters. It iterates through the real-time SOC values ​​of all battery clusters, sorts them by value to generate an ordered sequence, and extracts the maximum and minimum values ​​from the beginning and end of the sequence to calculate the range as the statistical dispersion. This range is compared with a preset dispersion threshold; if the threshold is exceeded, a balancing control mechanism is triggered.

[0030] Using the arithmetic mean of all state-of-charge (POC) values ​​as the dynamic equilibrium benchmark, the algebraic difference between each POC value and this benchmark is calculated to obtain the individual deviation. To eliminate the influence of system-level deviations, the average of all individual deviations is calculated as a common offset. Subtracting this common offset from each individual deviation yields the normalized power bias. Each power bias is encoded and integrated according to the address of its corresponding energy storage converter, forming a personalized power bias set. This bias set is mapped to the actual power adjustment range using weighting coefficients to ensure that the bias matches the system's power distribution capability.

[0031] During the process, the dispersion changes are continuously monitored, and when it falls below a threshold, the bias set is automatically cleared and the system exits the equalization mode. The energy management system temporarily stores the bias set in a buffer via a communication line, providing adjustment parameters for subsequent power allocation. The battery cluster operating data connected to the DC power line and the power commands transmitted via the AC power line are coupled at this step, laying the foundation for system-level optimization.

[0032] Step S40: The energy management system continuously acquires real-time state of charge (SOC) values ​​of battery clusters from each energy storage converter via communication lines, forming a monitoring closed loop. The monitoring data is based on a physical architecture where battery clusters are connected to the energy storage converters via DC power lines, and the energy storage converters are connected to the mains power via AC power lines. When the energy management system detects that the statistical dispersion of SOC values ​​among battery clusters exceeds a preset threshold, it triggers a balancing control mechanism.

[0033] Using the arithmetic mean of all state-of-charge values ​​as the equilibrium benchmark, the individual deviation of each battery cluster is calculated and integrated into a personalized power bias set. This bias set is used to adjust the power allocation strategy. When responding to grid power dispatch demands, the energy management system initially allocates the total power command according to the proportion of the remaining available capacity of each battery cluster, and then adds the bias to generate the desired power command.

[0034] During the execution phase, the energy management system sends independent control commands to each energy storage converter via communication lines. The energy storage converter adjusts its power conversion according to the commands: for battery clusters with a high state of charge (SOC), the command favors discharging or reduces charging power; for battery clusters with a low SOC, it favors charging or increases charging power. Charging and discharging energy is transferred between the battery clusters and the energy storage converters via DC power lines. The energy storage converters interact with the mains power supply via AC power lines, enabling applications such as peak shaving, valley filling, and frequency regulation.

[0035] The energy management system continuously monitors changes in the state of charge (SOC) value and updates the bias set in real time. When the statistical dispersion drops below a threshold, the system clears the bias set, exits the equalization mode, and enters normal operation. This cycle achieves data acquisition and command issuance through communication lines, with DC power lines transmitting battery energy and AC power lines completing grid coupling, forming a dynamic adaptive equalization to ensure system safety and economy.

[0036] This application provides a multi-type cascaded battery management method. By establishing an independent communication link for each battery cluster and implementing "one cluster, one management," it enables refined monitoring of heterogeneous cascaded batteries, laying the foundation for subsequent personalized energy dispatch and effectively solving the problem of coarse management caused by inconsistent battery degradation characteristics. By calculating the statistical dispersion of the state of charge (SOC) value and using it as the equalization trigger condition, combined with individual deviations to calculate the power bias, dynamic quantitative assessment and proactive intervention of consistency among battery clusters are achieved, thereby improving the overall available capacity and equalization efficiency of the system. By responding to grid dispatch requirements and integrating real-time status and power bias sets for optimized power allocation, it ensures that internal equalization strategies are executed while meeting external power demands, thereby improving the economy and response accuracy of the energy storage system when participating in peak-valley arbitrage, frequency regulation, and other applications. By issuing personalized power control commands to each independent energy storage converter, independent closed-loop control of the charging and discharging process of each battery cluster is achieved, effectively avoiding safety risks caused by overcharging or over-discharging of individual battery clusters and enhancing the overall reliability of the system.

[0037] In one embodiment, the energy management system sequentially polls the battery clusters connected to each energy storage converter using a communication protocol to establish an independent communication link with each battery cluster, including: The energy management system initiates a communication protocol polling process for the target battery cluster. A pre-built communication protocol library is stored in the energy management system's non-volatile memory, containing various industry-standard communication protocol specifications, such as CAN2.0B, Modbus-RTU, and TCP / IP. Each protocol record in the library includes frame structure definitions, baud rate parameters, verification algorithms, and timeout settings.

[0038] The polling process proceeds according to the preset priority order in the protocol library, starting with the most frequently used protocol types. The energy management system reads the parameter configuration of the currently tested protocol from the protocol library and assembles a handshake command frame according to the frame structure specification of that protocol. The assembly of the handshake command frame includes setting the synchronization header, filling the address field, selecting the function code, and constructing the data field.

[0039] The data field contains basic query commands, such as commands to read the battery's total voltage or SOC value. After frame assembly, a checksum is added; the CAN protocol uses CRC checksum, and the Modbus protocol uses LRC or CRC checksum. The energy management system sends the complete handshake command frame to the target energy storage converter via the communication line. The communication line uses shielded twisted-pair cable or fiber optic cable, and the physical layer conforms to RS-485 or Ethernet standards. After receiving the frame, the energy storage converter forwards it to the connected battery cluster, where the BMS within the battery cluster parses the command. A transmission timeout protection mechanism is implemented throughout the transmission process to ensure timely termination of the current session in case of link failure.

[0040] After receiving the handshake command frame, the energy storage converter transmits the command to the battery cluster's BMS system via the DC power line. The BMS parses the command according to its supported protocol specifications and generates a response data frame. The response data frame contains the battery cluster's status parameters, such as total voltage, total current, and SOC value.

[0041] The energy storage converter returns the response data from the BMS to the energy management system via a communication line. After sending the handshake command, the energy management system starts a timer to monitor the response timeout period. This period is dynamically adjusted according to the protocol type: 100-500ms for CAN protocol and 200-1000ms for Modbus protocol. If response data is received within the timeout period, a frame integrity check is performed to verify whether the frame header and trailer identifiers conform to the current trial protocol specifications. Next, the checksum field in the frame is extracted, and the data field is recalculated using the same checksum algorithm as the handshake command. The calculated checksum is compared with the checksum carried in the frame; if the comparison fails, a transmission error is determined.

[0042] During the verification process, the system simultaneously verifies whether the data field length conforms to the protocol specifications; data with abnormal lengths will be considered an invalid response. For response data that passes verification, the system records the received timestamp and signal quality indicators to provide a reference for subsequent protocol optimization. If verification fails or the response times out, the energy management system marks the current trial protocol as mismatched and prepares to switch to the next protocol for further attempts.

[0043] After the response data is verified by the checksum, the energy management system parses the data according to the frame structure definition of the current trial protocol. The parsing process extracts battery cluster status parameters from the response data frame according to the byte order and field length specified in the protocol. Typical status parameters include total battery cluster voltage, total current, SOC value, SOH value, temperature parameters, and alarm status words. The total voltage parameter is extracted from a specified byte position in the data frame; converting it to an engineering value requires multiplying it by a scaling factor defined in the protocol.

[0044] When parsing the total current parameter, the charging and discharging directions are distinguished; the discharging current is defined as positive, and the charging current as negative. State of Charge (SOC) and State of Harshness (SOH) values ​​are typically stored as percentages and converted to a 0-100% range after parsing. Temperature parameters may include multiple cell temperature values; the maximum value is used as the representative value. Alarm status words are parsed bit by bit, with each bit corresponding to a fault type such as overvoltage, undervoltage, or overtemperature.

[0045] Each state parameter obtained from the analysis is compared with a preset state value range. The voltage range is set according to the battery type; the allowable range for lithium iron phosphate battery clusters is typically 2.5V-3.65V multiplied by the number of series connections. The current range is determined based on the PCS rated power and battery capacity, with the upper limit not exceeding the maximum allowable charge / discharge current. The SOC range is fixed at 0-100%, and the SOH range is set at 70%-100%. The temperature range is set according to the cell specifications, with a typical value of -20℃ to 60℃. Parameter comparison uses an item-by-item verification method; any parameter exceeding the range is considered abnormal. During the comparison process, the specific values ​​and deviations of the parameters exceeding the limits are recorded to provide a reference for subsequent protocol selection. When all parameters are within a reasonable range, the state parameter set is marked as valid and proceeds to the next processing stage.

[0046] After all battery cluster status parameters pass range verification, the energy management system performs a protocol binding operation. The system generates a protocol configuration record, which includes the protocol type, baud rate parameters, frame format definition, and timeout settings. This record is associated with the physical address of the target energy storage converter and stored in the energy management system's configuration database.

[0047] During the binding process, communication parameters are set, configuring serial port parameters (data bits, stop bits, parity bits) or network parameters (IP address, port number) according to the protocol type. For the CAN protocol, an arbitration identifier and mask must be set; for the Modbus protocol, the device address and function code mapping relationship must be configured. After completing the software configuration, the system sends a test command to verify the stability of the communication link. The test command uses the standard query command of the protocol, sending it three times consecutively to verify the consistency of the response.

[0048] After the test is passed, the energy management system creates an independent data acquisition thread for the energy storage converter, sets the acquisition period and timeout retry mechanism. The physical basis of the independent communication link is a communication line connection. The energy management system is directly connected to the communication interface of the energy storage converter through the communication port, and the energy storage converter establishes communication with the BMS of the battery cluster through its internal circuitry.

[0049] Once the link is established, the energy management system can periodically acquire real-time data from the battery clusters and provide a communication channel for subsequent power control. If a communication interruption occurs during the bonding process, the system automatically switches to a backup protocol in the protocol library to re-attempt to establish a connection.

[0050] This embodiment automatically adapts to the communication protocols of different battery clusters by sequentially polling a pre-set communication protocol library and assembling corresponding handshake command frames. This effectively solves the communication protocol incompatibility problem caused by differences in battery batches and improves the system's compatibility with multiple types of cascaded batteries. By performing checksum verification and frame structure parsing on the response data, the integrity and accuracy of the communication data are ensured, providing a reliable data foundation for subsequent status monitoring. By comparing the parsed battery cluster status parameters with preset ranges one by one, abnormal battery cells can be quickly identified during the communication establishment phase, eliminating potential safety hazards in advance. By binding the verified protocol to the energy storage converter and establishing an independent communication link, precise monitoring of each battery cluster is achieved, laying the foundation for subsequent personalized power management.

[0051] In one embodiment, based on an independent communication link, the operating status data of the battery clusters uploaded by each energy storage converter are periodically collected, and the state of charge (SOC) is calculated to obtain a real-time SOC value, including: The energy management system sends data acquisition commands to each energy storage converter via an established independent communication link, according to a preset sampling period. Upon receiving the commands, the energy storage converter obtains operating status data from the connected battery clusters through its internal data acquisition module. The data acquisition module includes a high-precision current sensor and a real-time clock circuit, sampling the total charging and discharging current of the battery clusters at a fixed frequency to form a current sampling sequence. Each sampling point includes a current value and polarity indicator; discharging current is recorded as positive, and charging current as negative.

[0052] The timestamp sequence is generated by a high-precision timer, recording the precise time information of each current sampling point. The time interval is typically set to the millisecond level to ensure calculation accuracy. The energy storage converter encapsulates the current sampling sequence and timestamp sequence according to a predetermined data format, adds a frame header, device address, and checksum, and then transmits them to the energy management system via an independent communication link. After receiving the data, the energy management system performs an integrity check, verifying the data frame length, checksum matching, and timestamp continuity.

[0053] For data frames that fail verification, the energy management system records an error log and triggers a retransmission mechanism, requiring the energy storage converter to resend the operating status data for the current cycle. Data transmission via independent communication links ensures the isolation and real-time nature of the operating status data for each battery cluster, providing an accurate data foundation for subsequent state of charge calculations.

[0054] The energy management system preprocesses the received current sampling sequence and timestamp sequence, including outlier filtering and data alignment. The traversal process starts from the beginning of the sequence and processes each sampling point sequentially. For each current value in the current sampling sequence, the system reads the adjacent timestamps in the corresponding timestamp sequence and calculates the time interval between them as the sampling period.

[0055] The charge change within the sampling period is obtained by multiplying the current value by the time interval. The calculation process considers the current polarity; the charge change is negative during discharge and positive during charging. Data validity checks are performed during the calculation to identify and remove abnormal sampling points, such as sudden current changes or data with incorrect timestamps. For continuous sampling points, the trapezoidal integral method is used to improve calculation accuracy; the average current of two adjacent sampling points is multiplied by the time interval to obtain the charge change for that time period.

[0056] Each charge change result is temporarily stored in a buffer, labeled with its corresponding time interval and data type. After traversal, the system generates a charge change sequence that maintains the same time dimension as the original current sampling sequence. The charge change sequence undergoes smoothing filtering, and a moving average algorithm is used to eliminate random noise interference, ensuring data stability. The processed charge change sequence serves as intermediate calculation results, providing input data for subsequent cumulative calculations. The entire traversal calculation process is completed in the data processing unit of the energy management system, which is equipped with a high-precision floating-point arithmetic unit to ensure the accuracy and real-time performance of complex calculations.

[0057] The energy management system performs an accumulation operation on the sequence of charge changes. The accumulation process starts from the beginning of the sequence and algebraically sums the charge changes for each sampling period in chronological order. Negative charge changes during discharge and positive charge changes during charging are accumulated separately to obtain the final net charge change.

[0058] A data verification mechanism is implemented during the calculation process to identify and eliminate abnormal charge changes, such as abrupt changes caused by sampling interference. The rated capacity parameter is obtained from the battery cluster's technical specifications; this parameter represents the total usable capacity of the battery cluster under standard conditions. The system divides the total net charge change by the rated capacity parameter to obtain the cumulative charge change rate. A temperature correction factor is considered in the calculation process; the rated capacity value is compensated and adjusted based on the real-time temperature of the battery cluster to improve calculation accuracy.

[0059] The cumulative charge change rate represents the relative capacity change of the battery cluster since the last calibration; a positive value indicates net charging, and a negative value indicates net discharging. The calculation results undergo range validation to ensure they are within a reasonable range; outliers trigger a data recalculation mechanism. The cumulative charge change rate is temporarily stored as an intermediate variable in a register, providing input for the final state-of-charge (SOC) calculation.

[0060] The energy management system reads the initial state of charge (SOC) value of the battery cluster from the storage unit. This value is obtained from the battery management system during system initialization via a communication protocol. Algebraic superposition adds the cumulative rate of change of charge to the initial SOC value, and the superposition result serves as the preliminary calculation result for the real-time SOC value. A boundary protection mechanism is implemented during the calculation process; if the superposition result exceeds the standard range of 0-100%, it is automatically corrected to the nearest valid boundary value.

[0061] The calculation results are validated for reasonableness by comparing the trends of state of charge (SOC) changes over multiple consecutive cycles, identifying anomalous jumps, and triggering data verification. Verified real-time SOC values ​​are updated to the system database and simultaneously fed back to the energy storage converter via an independent communication link for cross-validation. The timestamps of the real-time SOC values ​​are synchronized with the current sampling sequence to ensure data timing consistency.

[0062] The calculated real-time state of charge (SOC) values ​​are used to update the system state dataset, providing data support for subsequent statistical dispersion calculations and power optimization allocation. The entire calculation process is completed within the data processing unit of the energy management system, with the SOC calculation for each battery cluster performed independently to ensure the system's parallel processing capabilities.

[0063] This embodiment acquires current sampling sequences and timestamp sequences through an independent communication link, ensuring the integrity and timing accuracy of the operating status data for each battery cluster, providing a reliable data foundation for subsequent state of charge (SOC) calculations. By traversing the sampling sequence and calculating charge changes, refined measurement of the battery charging and discharging process is achieved, effectively improving the accuracy of the cumulative charge change calculation. By accumulating charge changes and combining them with rated capacity parameters to calculate the cumulative charge change rate, the relative degree of battery capacity change can be accurately reflected, providing a scientific basis for SOC estimation. Algebraically superimposing the cumulative charge change rate with the initial SOC value enables real-time SOC estimation based on the ampere-hour integral method, ensuring the continuity and reliability of the calculation results.

[0064] In one embodiment, the statistical dispersion of the state of charge (SCC) values ​​is calculated. When the statistical dispersion exceeds a preset dispersion threshold, the individual deviation of each SCC value from a preset equilibrium benchmark is calculated. All deviations are then integrated to obtain a personalized power bias set, including: The energy management system obtains the latest state of charge (SOC) values ​​of all battery clusters in real time from each energy storage converter via established independent communication links. These values ​​are transmitted in the form of data packets, each containing a battery cluster identifier, SOC value, and timestamp. The system establishes a temporary data buffer to store the SOC values ​​of each battery cluster in the order they are received, and records the latest update time.

[0065] The processing begins with data validity verification, removing outliers outside the reasonable range (0%-100%) and marking battery clusters that have not been updated for multiple consecutive periods as invalid. After data collection is complete, the system sorts all valid state of charge (POC) values. A modified bubble sort algorithm is used to generate an ordered sequence by arranging the values ​​in ascending order. The first element of the ordered sequence is extracted as the lowest POC value, and the last element is extracted as the highest POC value.

[0066] The arithmetic difference between the highest and lowest values ​​is calculated; this difference represents the statistical dispersion, which characterizes the consistency level among battery clusters. An outlier detection mechanism is incorporated into the calculation process; when an extreme outlier is detected, a data verification process is initiated. The statistical dispersion calculation module simultaneously performs data sample size verification to ensure that the number of battery clusters involved in the calculation meets the minimum statistical requirements.

[0067] When the number of valid data points is insufficient, the system delays the calculation of dispersion and waits for more data input. The final statistical dispersion value is timestamped and stored in the system database for subsequent comparisons. The entire process is completed in the data preprocessing unit of the energy management system, which is equipped with a dedicated buffer and coprocessor to ensure real-time computing performance.

[0068] The decision logic unit of the energy management system calls a preset discrete threshold parameter, which is configured according to the battery type and application scenario, with a typical value of 5%-10%. The comparator module compares the statistical dispersion calculated in real time with this threshold to generate a binary judgment result.

[0069] When the statistical dispersion exceeds a threshold, the equalization control enable signal is triggered, initiating the individual deviation calculation process. The preset equalization benchmark is generated dynamically, taking the arithmetic mean of all current effective state of charge values. The calculation process uses a floating-point arithmetic unit to accumulate points and then calculate the average value, rounding the result to two decimal places.

[0070] The system iterates through the state of charge (SOC) value of each battery cluster, performing algebraic subtraction with the balancing benchmark in sequence. The difference is the individual deviation. The calculation process distinguishes between positive and negative polarities; deviations above the benchmark are positive, and those below are negative. The individual deviation calculation module implements a real-time verification mechanism to check the range of the calculation results. When the absolute value of the deviation exceeds a preset limit, the data recalculation process is initiated and the abnormal event is recorded. All calculated individual deviations are indexed and stored by battery cluster identifier, forming a temporary deviation set.

[0071] The system updates the historical deviation database, recording the changing trends of deviations for each battery cluster, providing data support for subsequent analysis. The processing is completed within the system's arithmetic logic unit, which has parallel processing capabilities and can simultaneously handle deviation calculations for multiple battery clusters.

[0072] The energy management system performs an arithmetic average on the set of individual deviations obtained in step two. The averaging process uses a weighted average algorithm, setting weight coefficients based on the rated capacity of each battery cluster. Battery clusters with larger capacities have higher weights in the average calculation.

[0073] A data verification mechanism is implemented during the calculation process to eliminate invalid deviations caused by communication anomalies, ensuring the accuracy of the average value. The resulting arithmetic mean is used as the common offset, which reflects the overall deviation of the system's state of charge. The individual deviations of each battery cluster are iterated and algebraically subtracted from the common offset. The calculation process distinguishes between positive and negative polarities: individual deviations higher than the common offset are subtracted from the offset to obtain a positive offset, while those lower than the common offset yield a negative offset. This calculation eliminates the influence of system-level deviations on individual battery clusters, making the offset more reflective of relative differences.

[0074] The result of each subtraction operation is a single power bias for that battery cluster. Its value range is normalized and limited to a preset power regulation range. All single power biases are organized and integrated according to their corresponding energy storage converter addresses to form a structured data set. Data consistency checks are performed during the integration process to ensure that each power bias corresponds to the correct battery cluster identifier.

[0075] The generated personalized power bias set contains complete metadata information, such as generation timestamp, data version number, and checksum. This bias set is stored in the energy management system's cache and backed up to non-volatile memory to prevent accidental data loss.

[0076] The system sets an expiration period for the bias set, which automatically expires and triggers a recalculation process after the expiration. The integrated personalized power bias set is transmitted to the power distribution module via the internal data bus, providing adjustment parameters for subsequent power optimization control.

[0077] This embodiment accurately quantifies the consistency differences between battery clusters by aggregating the real-time state of charge (SOC) values ​​of all battery clusters and calculating the range between the highest and lowest values ​​as the statistical dispersion. This provides a reliable triggering basis for equalization control and improves the system's accuracy in judging the consistency state. By comparing the statistical dispersion with a preset threshold and calculating the algebraic difference between each SOC value and the equalization benchmark, precise quantification of the deviation degree of each battery cluster is achieved, providing a data foundation for personalized power adjustment. By calculating the arithmetic mean of all individual deviations as a common offset and subtracting this offset from each individual deviation, the influence of system-level deviations on individual battery clusters is effectively eliminated, making the power offset more reflective of relative differences.

[0078] In one embodiment, the real-time state of charge (SOC) values ​​of all battery clusters are aggregated, and the difference between the highest and lowest SOC values ​​is extracted and calculated. This difference is used as the statistical dispersion, including: The energy management system acquires real-time state-of-charge (POC) data packets uploaded by each energy storage converter through established independent communication links. The data packet parsing module extracts the POC fields and corresponding battery cluster identifiers to establish a temporary dataset. The data processing unit initiates a traversal program, accessing each POC record in the dataset one by one.

[0079] During the traversal, data validity is validated, checking whether the values ​​are within a reasonable range (0%-100%) and verifying the freshness of the timestamps to exclude expired data. A modified bubble sort algorithm is used to achieve descending order. Starting from the first element of the dataset, the algorithm compares adjacent state-of-charge values ​​sequentially. When a value is found to be smaller than a subsequent value, the two values ​​are swapped. This process is repeated until the entire dataset is traversed, ensuring the maximum value is moved to the front of the sequence. This sorting process involves multiple iterations, each iteration reducing the number of comparison elements until all elements are arranged.

[0080] During the sorting process, data consistency checks are performed synchronously to identify and record abnormal jump values. Data with inconsistent sorting results across multiple consecutive sorting iterations is marked as pending verification, and a data retransmission mechanism is initiated. After sorting, an ordered state-of-charge (POC) sequence is generated and stored in a dedicated buffer. Each element in the sequence contains a POC value and its corresponding battery cluster identifier. The time complexity of the sorting algorithm has been optimized to ensure real-time performance in large-scale battery cluster scenarios. The generated ordered sequence provides a structured data foundation for subsequent extreme value extraction and discreteness calculations.

[0081] The data processing unit accesses the generated ordered state of charge (POC) sequence and locates the first element using the sequence pointer. The extraction operation first reads the data record at sequence index 0, parsing its POC value field as the highest POC value. Simultaneously, it obtains the corresponding battery cluster identifier and timestamp information to form a complete data record. Similarly, it accesses the end index position (n-1, where n is the sequence length), extracts the POC value of the last element as the lowest POC value, and records the corresponding metadata.

[0082] The extreme value extraction process includes a data verification step to check the logical rationality of the highest and lowest values. Verification includes checking numerical non-negativity, numerical range validation, and timestamp consistency confirmation. In anomalies, such as the simultaneous occurrence of both the highest and lowest values ​​for the same battery cluster, a data re-acquisition process is triggered. Extracted extreme value data is temporarily stored in a register, marked with the extraction time and data version information. Algebraic subtraction is performed using a fixed-point arithmetic unit, with the highest state of charge (SOC) value as the minuend and the lowest SOC value as the subtrahend. Considering numerical precision requirements, the results are rounded to two decimal places.

[0083] The calculation process employs an overflow protection mechanism, initiating an exception handling procedure when the result exceeds a preset range. The obtained absolute difference is normalized to eliminate the influence of dimensions, ensuring the result can be directly used for dispersion evaluation. The calculation result, along with the corresponding extreme value data, is stored in an intermediate result buffer to provide input data for subsequent statistical verification. The entire extraction and calculation process is completed in the arithmetic logic unit of the data processing unit, which is equipped with error detection and correction circuitry to ensure computational reliability.

[0084] The energy management system's data verification module accesses the length attribute of the ordered charge state sequence to obtain the number of valid data points currently participating in the calculation. This number is compared with a preset minimum statistical sample requirement, which is dynamically configured according to the system size and is typically set as a specific proportion of the total number of battery clusters.

[0085] The verification process uses a numerical comparator. The verification result is considered passed when the actual number of data points is greater than or equal to the minimum requirement. The data point quantity verification includes multiple checks. In addition to basic quantity comparison, it also verifies the timeliness distribution of the data points, ensuring that the timestamps of all data points are within a reasonable time window. Data points with outdated timestamps, even if the quantity requirement is met, will be marked as invalid and excluded from the statistical sample.

[0086] The system checks the source distribution of data points to ensure that the battery clusters involved in the statistics cover all online operating units, avoiding statistical distortion caused by communication interruptions in some units. When the verification conditions are met, the system officially outputs the absolute difference calculated in step two as the statistical dispersion. The output process includes data formatting operations, converting the absolute difference into a standardized dispersion representation and attaching a data quality identifier. The statistical dispersion value is assigned a timestamp and version number and stored in the dispersion record table of the system database. Simultaneously, this value is transmitted to the decision logic unit via the data bus as a quantitative indicator of the battery cluster consistency status. For cases where verification fails, the system initiates an exception handling procedure.

[0087] When the number of data points is insufficient, the dispersion calculation is delayed, and the system waits for new data in subsequent data collection cycles. Simultaneously, verification failure events are recorded, and the system status log is updated. If verification fails for multiple consecutive cycles, an alarm signal is triggered, and maintenance personnel are notified to check the communication link status. The entire verification process is completed in the system's quality control unit, which is equipped with a real-time monitoring mechanism to ensure the reliability and validity of the statistical dispersion output.

[0088] The decision logic unit of the energy management system calls a preset discrete threshold parameter, which is configured based on battery type, system operating mode, and historical data statistical analysis results. The comparator module compares the statistical dispersion output from step three with this threshold to generate a binary judgment result.

[0089] The comparison process employs a comparison algorithm with hysteresis characteristics to avoid frequent state jumps near the threshold. When the statistical dispersion exceeds the preset threshold, the comparator outputs a high-level signal, triggering the battery cluster equalization control. This enable signal activates subsequent processing steps, including modules for individual deviation calculation and power bias generation.

[0090] Simultaneously, relevant parameters of the triggered event are recorded, including the statistical dispersion value at the time of triggering, the timestamp, and the number of battery clusters involved in the calculation, forming a complete operation log. A real-time monitoring mechanism is implemented during the comparison and judgment process to continuously track the changing trend of the statistical dispersion. When the dispersion exceeds the threshold for multiple consecutive cycles, the equalization control parameters are automatically adjusted to increase the adjustment intensity. The rate of change of dispersion is also monitored; when a rapid increase trend is observed, an early warning is issued, and preventative control measures are initiated. The comparison results are distributed to relevant functional modules via the system's internal communication bus. When the dispersion exceeds the threshold, an equalization adjustment command is sent to the power distribution module; when the dispersion is below the threshold, a normal operation command is sent.

[0091] The entire comparison and judgment process is completed in the system's control and decision-making unit, which features a fault-safe design to ensure the system maintains a safe operating state under abnormal conditions. The comparison results are also used to update system status indicator lights, providing maintenance personnel with an intuitive display of the system's operating status.

[0092] This embodiment systematically organizes data and identifies extreme value distributions by traversing the real-time state of charge (SOC) values ​​of all battery clusters and sorting them by size to form an ordered sequence. This provides a reliable data foundation for subsequent dispersion calculations, improving the standardization and accuracy of data processing. Extracting the first and last elements directly from the ordered sequence as the highest and lowest SOC values ​​simplifies the extreme value acquisition process, avoids resource consumption from complex calculations, and improves system real-time performance. Calculating the absolute difference between extreme values ​​using algebraic subtraction directly reflects the maximum consistency deviation between battery clusters, providing a clear quantitative basis for equilibrium judgment. Verifying whether the number of data points meets the minimum statistical sample requirement ensures that dispersion calculations are based on sufficient data support, avoiding misjudgments due to insufficient samples and improving decision reliability. Outputting the absolute difference that meets the conditions as statistical dispersion achieves a standardized quantitative evaluation of the system's consistency state, providing an accurate trigger signal for subsequent power bias calculations, ultimately improving the accuracy of battery cluster collaborative management.

[0093] In one embodiment, in response to the power dispatching requirements of the mains power, power optimization allocation is performed by combining all real-time state of charge values ​​and personalized power bias sets, generating and issuing independent control commands to each energy storage converter to complete SOC balance management, including: The energy management system acquires real-time state-of-charge (POC) data packets uploaded by each energy storage converter via an independent communication link. The data packet parsing module extracts the POC field and the corresponding battery cluster identifier. The system accesses the configuration database and queries the corresponding rated power parameter based on the battery cluster identifier. This parameter includes two dimensions: maximum allowable charging power and maximum allowable discharging power.

[0094] The rated power parameters are derived from the battery cluster's technical specifications and are pre-stored in the database during system initialization, with a mapping relationship established between them and the battery cluster identifier. The remaining available capacity calculation module initiates the calculation process, reading the battery cluster's rated capacity parameters and the current real-time state of charge (SOC) value. The calculation process is executed using a fixed-point arithmetic unit, converting the real-time SOC value into decimal form before multiplying it by the rated capacity.

[0095] The calculation process considers a temperature compensation coefficient and corrects the rated capacity value based on the real-time temperature of the battery cluster to improve calculation accuracy. The remaining available capacity result is stored in ampere-hours (AH) and simultaneously converted to a percentage for subsequent proportional calculations. The data verification stage performs a range check on the calculation results to ensure that the remaining available capacity is within a reasonable range. For abnormal calculation results, the system initiates a data review mechanism, re-acquiring real-time state-of-charge values ​​for secondary calculation. The calculated remaining available capacity value, together with the rated power parameters of the corresponding battery cluster, forms a temporary data set, which is timestamped and stored in a cache. The entire processing is completed collaboratively by the system's data query module and arithmetic logic unit, ensuring the real-time performance and accuracy of data acquisition and calculation.

[0096] The power allocation module of the energy management system reads the temporary data set generated in step one and extracts the remaining available capacity and rated power parameters of each battery cluster. Power ratio calculation is implemented using a division operator, dividing the remaining available capacity by the rated power parameter to obtain a standardized ratio coefficient. The calculation process includes a non-zero division check; if the rated power parameter is zero or an outlier, a backup parameter is used for alternative calculation. The ratio coefficients are normalized by summing the ratio coefficients of all battery clusters and calculating the proportion of each battery cluster's coefficient in the sum. This proportion is the power allocation weight of each battery cluster, with the weight value controlled between 0 and 1 to ensure that the sum of all weights equals 1. After the weight calculation is completed, the system accesses the power dispatch demand buffer and reads the total power command value for the mains power dispatch.

[0097] The initial allocation process employs a weighted allocation algorithm, multiplying the total power command value by the power allocation weight of each battery cluster. The multiplication operation uses a high-precision multiplier, retaining the result to two decimal places. The allocation process distinguishes between power directions: charging commands are allocated negative power, and discharging commands are allocated positive power. The base power value allocated to each battery cluster is temporarily stored in a result register, marked with an allocation timestamp and version number. A data verification mechanism performs a total verification of the allocation results, ensuring that the error between the sum of all base power values ​​and the total power command value is within acceptable limits.

[0098] When the error exceeds the limit, a proportional adjustment algorithm is activated to reallocate the power. The final set of base power values ​​is transmitted to the next processing stage via the data bus, and simultaneously backed up to non-volatile memory to prevent data loss. The entire calculation process is completed in the power management unit, which is equipped with a dedicated coprocessor to ensure real-time performance for complex calculations.

[0099] The power adjustment module of the energy management system accesses a personalized power bias set storage area, which contains power adjustment value records for each battery cluster. The data extraction process uses index matching based on the battery cluster identifier to ensure accurate correspondence between each base power value and its corresponding power adjustment value. The power adjustment value is stored as a percentage, representing the relative adjustment magnitude based on the base power value.

[0100] The power adjustment calculation is implemented using algebraic superposition, performing an algebraic addition operation between the base power value and the power adjustment value. The adjustment value has positive and negative polarities; a positive value indicates an increase in output power, and a negative value indicates a decrease in output power. The calculation process considers the consistency of power direction: in charging mode, a negative adjustment indicates an increase in charging power, and in discharging mode, a positive adjustment indicates an increase in discharging power. The calculation is performed using a high-precision arithmetic logic unit, retaining two decimal places of precision. The adjusted expected power value undergoes a range pre-check to identify abnormal results that may exceed safety limits.

[0101] For any outliers detected during the pre-check, the system initiates an adjustment value scaling mechanism, proportionally reducing the range of all adjustment values ​​to ensure the expected power value remains within a reasonable range. The calculated set of expected power values ​​is assigned a version identifier and timestamp and stored in a temporary result buffer. The data consistency verification step checks the integrity of the expected power value set, ensuring that each battery cluster has a corresponding calculation result. For missing data, the system uses the previously valid adjustment value for interpolation calculation. The verified set of expected power values ​​is transmitted to the limiting processing module via the internal data bus, providing input data for subsequent processing steps.

[0102] The limiting module reads the rated power parameters of each battery cluster from the system database, including the maximum allowable charging power and the maximum allowable discharging power limits. Version checks are performed during parameter acquisition to ensure that the rated power parameters used match the actual configuration of the current battery cluster. The system establishes an independent limiting comparator for each battery cluster to handle limiting judgments in both charging and discharging power directions. A bidirectional limiting algorithm is used for limiting the desired power value.

[0103] For discharge power, the desired power value is compared with the maximum permissible discharge power. When the desired power value exceeds the limit, the target power value is limited to the maximum permissible discharge power value. For charging power, the absolute value of the desired power value is compared with the maximum permissible charging power. When the limit is exceeded, the negative value of the maximum permissible charging power is taken as the target power value. The limiting process implements a soft limiting strategy. When the power value is detected to be close to the limit, a smooth transition algorithm is automatically activated to avoid sudden power changes.

[0104] The processed power value is rounded to the minimum power adjustment accuracy unit specified by the system. The target output power values ​​of all battery clusters are organized in a uniform format to generate a target output power value set. The set data undergoes integrity verification to check whether the target power value for each battery cluster has been calculated. The verified target output power value set is marked as ready and stored in the output buffer to await invocation by the instruction generation module. Simultaneously, the system records constraint triggering during the limiting process for subsequent system performance analysis and optimization.

[0105] The instruction generation module reads the target output power value set and assembles power control instruction frames according to a predefined communication protocol format. Each instruction frame contains fields such as the target power value, power direction identifier, and execution timestamp. The frame structure is adapted to different energy storage converter communication protocols. A unique sequence number and checksum are added to each instruction frame to ensure transmission integrity and traceability. The instruction issuance process employs a time-division multiplexing mechanism, sending control instructions sequentially to each energy storage converter via independent communication links.

[0106] The transmission timing is optimized based on the response characteristics and communication latency of the energy storage converter to ensure synchronous command execution. A handshake confirmation mechanism is implemented during transmission, waiting for the energy storage converter to return a reception confirmation signal before sending the next command. After receiving the power control command, the energy storage converter parses the command content and converts it into drive signals for the power semiconductor devices.

[0107] The drive signal generation takes into account the real-time status of the battery clusters and employs a closed-loop control method to ensure that the actual output power matches the target value. During execution, the energy storage converter monitors the operating parameters of the battery clusters in real time and automatically activates the protection mechanism when an anomaly is detected. The energy management system continuously monitors the power execution and state of charge (SOC) changes of each battery cluster and evaluates the SOC balancing effect through statistical dispersion indices.

[0108] When the statistical dispersion remains below a preset threshold, the SOC equalization management is considered complete. The system then clears the personalized power bias set and enters normal operation mode. The execution status and performance evaluation data of the entire process are recorded in the system log for subsequent performance analysis and optimization reference.

[0109] This embodiment obtains rated power parameters based on real-time state of charge (SOC) values ​​and calculates remaining available capacity, accurately reflecting the real-time energy state of each battery cluster. This provides a scientific basis for power allocation and improves the rationality of power scheduling. By calculating the ratio of remaining available capacity to rated power parameters and performing initial allocation, it ensures that power allocation matches the actual capacity of each battery cluster, avoiding overload or underload situations. By extracting adjustment values ​​from the power bias set to calibrate the base power, it achieves active balancing between battery clusters while meeting grid dispatch requirements, improving consistency regulation. Limiting the desired power value based on rated power parameters effectively prevents power commands from exceeding the safe operating range of battery clusters, ensuring system reliability. By generating and executing independent power control commands, precise control of each battery cluster is achieved, ultimately completing the closed-loop operation of SOC balancing management and improving overall system performance.

[0110] In one embodiment, limiting the desired power value based on the rated power parameter to obtain the target output power value includes: The parameter extraction module of the energy management system accesses the configuration database and queries the corresponding rated power parameter record based on the battery cluster identifier. The rated power parameter includes two dimensions: maximum allowable charging power and maximum allowable discharging power. These parameters are imported from the battery technical specifications during the system initialization phase and a permanent mapping relationship is established with each battery cluster.

[0111] The query process employs a version control mechanism to ensure that the parameters used are a perfect match for the current battery cluster hardware configuration. Expected power values ​​are read from a temporary result buffer, which stores intermediate results processed by the power adjustment module. Each expected power value carries a battery cluster identifier and timestamp information to ensure accurate correspondence with the rated power parameters. The data matching process uses an identifier comparison algorithm to associate and match expected power value records with rated power parameter records.

[0112] The comparison operation is performed by a dedicated numerical comparator, which differentiates the comparison process based on power direction. For a positive desired power value, it is compared to the maximum permissible discharge power parameter; for a negative desired power value, its absolute value is compared to the maximum permissible charging power parameter. The comparator outputs three states: exceeding the upper limit, below the lower limit, or within the normal range.

[0113] A real-time monitoring mechanism is implemented during the comparison process, and an early warning flag is activated when the expected power value is detected to be close to the limit. The comparison result, along with a timestamp and version number, is stored in an intermediate register to provide a basis for subsequent limiting processing. The entire comparison process is completed in the system's safety control unit, which is equipped with redundant verification circuitry to ensure the reliability of the comparison results.

[0114] The limiting processing module reads the comparator's output status and performs corresponding assignment operations based on different status branches. For cases exceeding the upper limit, the system extracts the corresponding upper limit value from the rated power parameter record and uses it as the target output power value.

[0115] The assignment process considers the consistency of power direction, using positive values ​​for discharge power and negative values ​​for charging power to ensure the correctness of the physical meaning. For cases below the lower limit, the processing module extracts the lower limit value of the rated power parameter as the target output power value. The assignment of the lower limit value also distinguishes between power direction, using positive values ​​for discharge power and negative values ​​for charging power. Boundary checks are performed during the processing to ensure that the assignment results comply with system safety specifications. When the desired power value is within the normal range, the system directly copies the desired power value as the target output power value.

[0116] The replication process performs data integrity checks, verifying the consistency of the source and target data formats. All assignment operations are logged, including information such as the original expected power value, limiting type, and final target value. After assignment, the target output power value is normalized to the minimum power adjustment accuracy unit specified by the system.

[0117] The processing results are stored in the target power value buffer and marked with a processing status flag. The target output power value of each battery cluster is stored independently, maintaining the original identifier mapping relationship, in preparation for subsequent aggregation.

[0118] The data aggregation module reads the processing results of all battery clusters from the target power value buffer and collects data in the order of battery cluster identifiers. The collection process performs an integrity check to verify that each battery cluster has a corresponding target output power value record. For missing data, the system initiates a data completion mechanism, using the previously valid value for interpolation or triggering a recalculation process.

[0119] The aggregation operation employs a structured data organization method, encapsulating the target output power value of each battery cluster along with its corresponding identifier, timestamp, and limiting status flag. Data encapsulation follows a predefined communication protocol format, ensuring that subsequent instruction generation modules can directly parse and use it. Data alignment is performed during encapsulation, converting all values ​​to standard units and precision formats.

[0120] The resulting set of target output power values ​​undergoes multiple checks, including summation check, range check, and logical relationship check. The summation check verifies whether the sum of all target power values ​​matches the total scheduled power requirement; if the error exceeds the allowable range, an automatic adjustment algorithm is activated. The range check confirms that each power value is within its corresponding safe operating range.

[0121] The logical relationship verification checks the match between the power direction and the battery cluster state to avoid logical conflicts. The set of target output power values ​​that pass the verification is marked as ready and stored in the system output buffer. The set data is also backed up to non-volatile memory to prevent accidental data loss. The system generates a version number and checksum for the set data, establishing proof of data integrity.

[0122] The aggregated target output power values ​​are transmitted to the command generation module via the internal data bus, triggering the subsequent power control command generation process. The entire aggregation process is completed in the system's data management unit, which is equipped with error detection and correction mechanisms to ensure the reliability of data processing. The time performance of the aggregation operation has been optimized to meet the requirements of real-time system control, providing an accurate and reliable data foundation for the final power command issuance.

[0123] This embodiment extracts the rated power parameters corresponding to each battery cluster and compares them with the expected power value, ensuring that the power command is always within the safe operating range, effectively preventing overload or underload situations, and improving the safety and reliability of the system. Through a conditional branching mechanism, target output power values ​​are set for different situations where the expected power value exceeds the upper limit, falls below the lower limit, or is within the normal range, achieving precise power limiting control and enhancing the system's adaptability and flexibility. Using the upper and lower limits of the rated power parameters as boundary conditions for limiting processing allows for personalized power constraints based on the actual capabilities of each battery cluster, avoiding performance losses caused by conventional limiting methods. By aggregating the target output power values ​​after limiting processing into a set, the matching of the system's total output power with scheduling requirements is ensured, while maintaining the coordination and consistency of power commands from each battery cluster. The entire limiting process, through multi-level verification and standardization operations, ensures the accuracy and usability of the target output power value set, providing a reliable data foundation for subsequent power control command generation.

[0124] Reference Figure 3 As shown, this application also provides a multi-type cascade battery management system, applicable to any of the above-mentioned multi-type cascade battery management methods, including: The data acquisition module is used to poll the battery clusters connected to each energy storage converter in turn through the energy management system to establish an independent communication link with each battery cluster. The analysis module is used to periodically collect the battery cluster operating status data uploaded by each energy storage converter based on an independent communication link, and to calculate the state of charge to obtain the real-time state of charge value. The correlation module is used to calculate the statistical dispersion of the state of charge values. When the statistical dispersion exceeds the preset dispersion threshold, the individual deviation of each state of charge value from the preset equilibrium benchmark is calculated, and all individual deviations are integrated to obtain a personalized power bias set. The processing module is used to respond to the power dispatching requirements of the mains power, combine all real-time state of charge values ​​and personalized power bias sets to optimize power allocation, generate and send independent control commands to each energy storage converter, and complete SOC balance management.

[0125] This application provides a multi-type cascaded battery management system. By establishing an independent communication link for each battery cluster and implementing "one cluster, one management," it enables refined monitoring of heterogeneous cascaded batteries, laying the foundation for subsequent personalized energy dispatch and effectively solving the problem of coarse management caused by inconsistent battery degradation characteristics. By calculating the statistical dispersion of the state of charge (SOC) value and using it as the equalization trigger condition, combined with individual deviations to calculate the power bias, it achieves dynamic quantitative assessment and proactive intervention of consistency among battery clusters, thereby improving the overall available capacity and equalization efficiency of the system. By responding to grid dispatch requirements and integrating real-time status and power bias sets for optimized power allocation, it ensures that internal equalization strategies are executed while meeting external power demands, thereby improving the economy and response accuracy of the energy storage system when participating in peak-valley arbitrage, frequency regulation, and other applications. By issuing personalized power control commands to each independent energy storage converter, it achieves independent closed-loop control of the charging and discharging process of each battery cluster, effectively avoiding safety risks caused by overcharging or over-discharging of individual battery clusters and enhancing the overall reliability of the system.

[0126] Reference Figure 4 As shown, this application also provides a multi-type cascade battery management device, including: Memory, used to store programs; A processor is used to execute programs to implement the various steps of a multi-type cascade battery management method, which includes any of the above-mentioned features.

[0127] This application also provides a storage medium storing computer instructions for causing a computer to perform any of the methods described above.

[0128] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments concerning the apparatus in the above embodiments, and will not be elaborated further here.

[0129] The solution of this application has been described in detail above with reference to the accompanying drawings. In the above embodiments, the descriptions of each embodiment have different emphases; parts not described in detail in a certain embodiment can be referred to in the relevant descriptions of other embodiments. Those skilled in the art should also understand that the actions and modules involved in the specification are not necessarily essential to this application. Furthermore, it is understood that the steps in the method of this application embodiment can be adjusted, combined, and deleted according to actual needs, and the modules in the device of this application embodiment can be combined, divided, and deleted according to actual needs.

[0130] Furthermore, the method according to this application can also be implemented as a computer program or computer program product, which includes computer program code instructions for performing some or all of the steps in the method described above.

[0131] Alternatively, this application may be implemented as a non-transitory machine-readable storage medium (or computer-readable storage medium, or machine-readable storage medium) storing executable code (or computer program, or computer instruction code) thereon, which, when executed by a processor of an electronic device (or electronic device, server, etc.), causes the processor to perform part or all of the steps of the methods described above according to this application.

[0132] Those skilled in the art will also understand that the various exemplary logic blocks, modules, circuits, and algorithm steps described in connection with the present application can be implemented as electronic hardware, computer software, or a combination of both.

[0133] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems and methods according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than those marked in the drawings. For example, two consecutive blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0134] The various embodiments of this application have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles, practical application, or improvement of the technology in the market, or to enable others skilled in the art to understand the embodiments disclosed herein.

Claims

1. A method for managing multiple types of cascaded batteries, characterized in that, include: The energy management system sequentially polls the battery clusters connected to each energy storage converter using the communication protocol to establish an independent communication link with each battery cluster. Based on the independent communication link, the battery cluster operation status data uploaded by each of the energy storage converters is periodically collected, and the state of charge is calculated to obtain the real-time state of charge value. Calculate the statistical dispersion of the state of charge values. When the statistical dispersion exceeds a preset dispersion threshold, calculate the individual deviation of each state of charge value from the preset equilibrium benchmark. Integrate all the individual deviations to obtain a personalized power bias set. In response to the power dispatching requirements of the mains power, power optimization allocation is performed by combining all the real-time state of charge values ​​and the personalized power bias set, generating and issuing independent control commands to each of the energy storage converters to complete SOC balance management; The step of polling the battery clusters connected to each energy storage converter sequentially using the communication protocol through the energy management system to establish an independent communication link with each battery cluster includes: For a given battery cluster, a protocol type is selected sequentially from a pre-set communication protocol library as the current trial protocol, assembled into a corresponding handshake instruction frame, and sent to the energy storage converter connected to the battery cluster. When the response data from the energy storage converter is received within the preset timeout period, the response data is verified according to the specifications of the current trial protocol. If the response data passes verification, the frame structure of the response data is parsed to obtain the battery cluster status parameters, and the battery cluster status parameters are compared one by one with the preset status value range. When all the state parameters of the battery cluster are within the range of the state values, it is determined that the communication protocol of the battery cluster is successfully matched, and the corresponding energy storage converter is bound to the current trial protocol to complete the establishment of the independent communication link.

2. The multi-type cascade battery management method according to claim 1, characterized in that, Based on the independent communication link, the system periodically collects the battery cluster operating status data uploaded by each of the energy storage converters, and performs state of charge calculation to obtain a real-time state of charge value, including: The battery cluster operating status data, including current sampling sequences and timestamp sequences, is obtained from each of the energy storage converters via the independent communication links. By traversing the current sampling sequence and the timestamp sequence, the current value of each sampling point in the current sampling sequence is multiplied by the sampling time interval in the timestamp sequence to obtain the charge change in each sampling period; The cumulative charge change rate is obtained by summing the charge changes over all the sampling periods and combining them with the rated capacity parameter of the battery cluster. The cumulative rate of change of charge is algebraically superimposed with the initial state of charge value of the battery cluster, and the superposition result is set as the real-time state of charge value.

3. The multi-type cascade battery management method according to claim 1, characterized in that, The statistical dispersion of the state of charge (SCC) values ​​is calculated. When the statistical dispersion exceeds a preset dispersion threshold, the individual deviation of each SCC value from a preset equilibrium benchmark is calculated. All individual deviations are then integrated to obtain a personalized power bias set, including: The real-time state of charge values ​​of all the battery clusters are collected, and the difference between the highest and lowest real-time state of charge values ​​is extracted and calculated. The difference is used as the statistical dispersion. The statistical dispersion is compared with the preset dispersion threshold. If the statistical dispersion is greater than the preset dispersion threshold, the algebraic difference between each state of charge value and the preset equilibrium benchmark is calculated, and the algebraic difference is set as the individual deviation. Calculate the arithmetic mean of all the individual deviations as a common offset, subtract the common offset from each individual deviation, and integrate the resulting individual power offsets to obtain the personalized power offset set.

4. The multi-type cascade battery management method according to claim 3, characterized in that, The process of aggregating the real-time state of charge (SOC) values ​​of all the battery clusters, extracting and calculating the difference between the highest and lowest SOC values, and using this difference as the statistical dispersion includes: The real-time state of charge values ​​of all the battery clusters are traversed, and the real-time state of charge values ​​are sorted in descending order to form an ordered state of charge sequence. Extract the first element from the ordered sequence of charged states as the highest charged state value, and extract the last element as the lowest charged state value. The absolute difference between the highest state of charge value and the lowest state of charge value is obtained by algebraically subtracting the two values. Verify whether the number of data points in the ordered charged state sequence meets the preset minimum statistical sample requirement. When the requirement is met, output the absolute difference as the statistical dispersion.

5. The multi-type cascade battery management method according to claim 1, characterized in that, The system responds to the power dispatching needs of the mains power supply, combines all the real-time state of charge values ​​with the personalized power bias set to perform power optimization allocation, generates and sends independent control commands to each of the energy storage converters, and completes SOC balance management, including: Based on the real-time state of charge value, the rated power parameters of the corresponding battery cluster are obtained, and the remaining usable capacity of the battery cluster is calculated. Calculate the power ratio of the remaining available capacity to the rated power parameter, and perform an initial allocation of the power scheduling demand based on the power ratio to obtain the base power value; Extract the power adjustment value from the personalized power bias set and adjust the base power value to obtain the desired power value; Based on the rated power parameters, the desired power value is limited to obtain a set of target output power values; Based on the target output power value set, a corresponding power control command is generated for each energy storage converter, and then sent to each energy storage converter for execution through the independent communication link to complete SOC equalization management.

6. The multi-type cascade battery management method according to claim 5, characterized in that, The step of limiting the desired power value based on the rated power parameters to obtain the target output power value includes: Extract the rated power parameter corresponding to each battery cluster and compare it with the expected power value respectively; When the desired power value exceeds the upper limit of the corresponding rated power parameter, the upper limit of the rated power parameter is set as the target output power value; When the desired power value is lower than the corresponding lower limit of the rated power parameter, the lower limit of the rated power parameter is set as the target output power value; When the desired power value is within the upper and lower limits of the corresponding rated power parameter, the desired power value is directly set as the target output power value; The target output power values ​​of all the battery clusters after the limiting process are summarized to form the target output power value set.

7. A multi-type cascade battery management system, characterized in that, The multi-type cascade battery management method applied to any one of claims 1-6 includes: The acquisition module is used to poll the battery clusters connected to each energy storage converter in turn through the energy management system to establish an independent communication link with each battery cluster. The analysis module is used to periodically collect the battery cluster operating status data uploaded by each of the energy storage converters based on the independent communication link, and to perform state of charge calculation to obtain real-time state of charge values. The association module is used to calculate the statistical dispersion of the state of charge value. When the statistical dispersion exceeds a preset dispersion threshold, the module calculates the individual deviation of each state of charge value from the preset equilibrium benchmark and integrates all the individual deviations to obtain a personalized power bias set. The processing module is used to respond to the power dispatching requirements of the mains power, combine all the real-time state of charge values ​​and the personalized power bias set to perform power optimization allocation, generate and send independent control commands to each of the energy storage converters, and complete SOC balance management. The step of polling the battery clusters connected to each energy storage converter sequentially using the communication protocol through the energy management system to establish an independent communication link with each battery cluster includes: For a given battery cluster, a protocol type is selected sequentially from a pre-set communication protocol library as the current trial protocol, assembled into a corresponding handshake instruction frame, and sent to the energy storage converter connected to the battery cluster. When the response data from the energy storage converter is received within the preset timeout period, the response data is verified according to the specifications of the current trial protocol. If the response data passes verification, the frame structure of the response data is parsed to obtain the battery cluster status parameters, and the battery cluster status parameters are compared one by one with the preset status value range. When all the state parameters of the battery cluster are within the range of the state values, it is determined that the communication protocol of the battery cluster is successfully matched, and the corresponding energy storage converter is bound to the current trial protocol to complete the establishment of the independent communication link.

8. A multi-type cascade battery management device, characterized in that, include: Memory, used to store programs; A processor is configured to execute the program to implement the various steps of the multi-type cascade battery management method as described in any one of claims 1-6.

9. A storage medium, characterized in that, The computer contains computer instructions for causing the computer to perform the method according to any one of claims 1 to 6.

Images