Intelligent battery hybrid parallel access system and control method

Abstract

The application relates to the technical field of intelligent battery management, and discloses an intelligent battery hybrid parallel access system and a control method. The method divides a battery pack into a basic energy storage unit and a dynamic adjustment unit, constructs a hybrid parallel access network based on a hierarchical interaction architecture, collects unit voltage, loop current and temperature field distribution data through a multi-dimensional sensing module, extracts features and fuses weights by using a time sequence correlation processor, inputs a state decision module to adjust an access mode, and executes mode switching by a double-path instruction parser. Meanwhile, real-time data and reference parameters are compared through a hierarchical evaluation mechanism, an operation deviation is determined, an adjustment signal is transmitted through a feedback channel after the operation deviation is determined, and an operation record is generated, and the response speed of the adjustment signal is positively correlated with the deviation. The method comprises a hybrid architecture construction module, a multi-source data processing module and an operation deviation feedback module. The application realizes efficient management and accurate control of the battery pack, and improves the flexibility, stability and safety of the system.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent battery management, and in particular to an intelligent battery hybrid parallel access system and a control method. Background Art

[0002] As energy storage systems continue to evolve, hybrid parallel battery access technology is becoming increasingly widespread. However, existing technologies face numerous challenges in practical operation. Traditional battery access control methods often lack a scientific and rational unit division mechanism, making it impossible to accurately classify battery cells based on their capacity characteristics and responsiveness. This leads to confusion between basic energy storage and dynamic regulation functions, making it difficult for the system to achieve efficient energy distribution and dynamic adjustment under different operating conditions.

[0003] Existing systems lack the ability to collect and process multi-source data. The collection of data such as unit voltage, loop current, and temperature field distribution lacks systematicity. The time-series correlation processing methods are simple, making it difficult to extract effective features from complex multi-source data and perform weighted fusion. This significantly reduces the accuracy and real-time nature of state decisions, making it impossible to adjust the access mode based on battery status in a timely manner. Furthermore, the evaluation and feedback mechanisms for operational deviations also have significant flaws. The comparison and analysis of real-time collected data with baseline parameters is not comprehensive enough, the feedback channel is single and has a slow response speed, and it is impossible to generate effective adjustment signals in a timely manner based on the degree of operational deviation. This leads to performance fluctuations in the battery pack during operation, even affecting the safety and stability of the system.

[0004] The regional division strategy lacks dynamic adjustment capabilities and cannot update the sub-region division in real time according to the changes in the battery pack's charge and discharge status and operating status. This makes the configuration priority of the control node unreasonable and the flexibility and adaptability of the overall access architecture poor. The semantic parsing model of the dual-path instruction parser is not perfect, and the validity judgment of voltage and current instructions is unreliable. The ability to handle abnormal signals is insufficient, which can easily lead to mode switching errors and affect the normal operation of the system. Summary of the Invention

[0005] The object of the present invention is to provide a smart battery hybrid parallel access system and control method to solve the problems raised in the above background technology.

[0006] To achieve the above objectives, the present invention provides a smart battery hybrid parallel access system and control method, the method comprising:

[0007] The battery pack is divided into basic energy storage units and dynamic adjustment units, and a hybrid parallel access network is constructed based on a hierarchical interactive architecture;

[0008] The multi-dimensional sensing module collects unit voltage, loop current and temperature field distribution data, and uses a time-series correlation processor to extract features and fuse weights from multi-source data. The fused data is input into the state decision module to adjust the access mode in real time, and the dual-path instruction parser executes the mode switching operation.

[0009] Through a hierarchical evaluation mechanism, the real-time collected data is compared and analyzed with the benchmark parameters to determine the operating deviation of the battery pack. The adjustment signal is transmitted to the control terminal through the feedback channel and an operation record is generated. The feedback channel includes a current compensation channel, a temperature adjustment channel and a status prompt channel. Among them, the response speed of the adjustment signal is positively correlated with the operating deviation.

[0010] Preferably, the layered interaction architecture constructs a hybrid parallel access network, comprising the following steps:

[0011] Set the classification criteria for unit functions, traverse each energy storage unit in the battery pack, evaluate its capacity characteristics and response capabilities, and record the functional attributes of each unit. At the same time, identify the units in the battery pack with changing charge and discharge states, and use a regional partitioning strategy to dynamically divide the battery pack into a basic energy storage area and a dynamic adjustment area;

[0012] Each functional area is assigned to a different control node. A new global synchronization identifier is generated according to the system operating frequency and sent to each control node. Each node performs access parameter configuration operations based on the assigned functional area and the current synchronization identifier, and integrates the configuration results of all functional areas into the main control network to form a complete hybrid parallel access architecture.

[0013] Preferably, the area division strategy includes the following steps:

[0014] The monitoring module tracks the operating status of the battery pack in real time, including charge and discharge power and temperature distribution. The core adjustment area is determined according to the operating status. Taking the core area as the benchmark, the unit combination within the set range is expanded outward, and the battery pack is recursively divided into sub-areas. The degree of correlation between the sub-area and the benchmark is used as the basis for configuration priority. Sub-areas with the same degree of correlation are coordinated through synchronization identification, and the sub-area division is dynamically updated according to changes in the operating status.

[0015] Preferably, the time series association processor performs feature extraction and weight fusion on multi-source data, including the following steps:

[0016] Real-time collection of unit voltage data, periodic collection of loop current data, and continuous collection of temperature field distribution data;

[0017] The different types of data are aligned in the time dimension by the clock synchronization module, and different weights are given according to the stability and influence degree of the data, a storage unit for processing historical data and a calculation unit for analyzing real-time data are designed to form a composite processing structure, multi-source data is input into the storage unit and the calculation unit, the feature information of the multi-source data is extracted, and the comprehensive state characteristics are generated by superimposed fusion according to the weight.

[0018] Preferably, the dual-channel instruction parser includes a voltage instruction channel and a current instruction channel, and the mode switching operation is performed by the dual-channel instruction parser, including the following steps:

[0019] The voltage instruction channel adopts a feature extraction and sequence decoding model for semantic analysis, the feature extraction module extracts voltage fluctuation features, the sequence decoding module analyzes the timing dependency, outputs a structured adjustment instruction, and sets a recognition confidence score for determining the validity of the voltage instruction, if the recognition confidence score is lower than the set threshold, an abnormal signal is output, if the recognition confidence score is not lower than the set threshold, a switching instruction is output;

[0020] The current instruction channel is based on a space correlation and time tracking model for semantic analysis, the continuous sampling data of the loop current is input, the correlation analysis module captures the correlation of spatial distribution and time variation, outputs a current adjustment code, and sets an analysis confidence score for determining the validity of the current instruction, if the analysis confidence score is lower than the set threshold, an abnormal signal is output, if the analysis confidence score is not lower than the set threshold, a switching instruction is output.

[0021] Preferably, the parser performs mode switching operation, including the following scenarios:

[0022] Scenario one, receiving collected unit voltage, loop current and temperature field distribution data, generating continuous sampling data of the loop current, triggering the current instruction channel to output the current adjustment code, and performing mode switching operation according to the switching instruction;

[0023] Scenario two, receiving real-time voltage fluctuation data, extracting voltage fluctuation features, triggering the voltage instruction channel to output a structured adjustment instruction, and performing mode switching operation according to the switching instruction;

[0024] Scenario three, sensing the abnormal signals output by the voltage instruction channel and the current instruction channel, if the number of abnormal signals exceeds the warning threshold, establishing a voltage-current instruction mapping table, weighting and fusing the confidence scores of the voltage channel and the current channel, outputting the matching degree of the dual-channel instruction and the adjustment strategy, and setting a comprehensive confidence, if the comprehensive confidence exceeds the total threshold, determining that the instruction is valid and triggering mode switching, otherwise, sending an alarm signal to the control terminal.

[0025] Preferably, determining the operating deviation of the battery pack comprises the following steps:

[0026] Establish baseline operating parameters as control targets, including a baseline voltage range, a baseline current threshold, and a baseline temperature range, and align the baseline operating parameters with the battery pack's real-time operating data in the time dimension;

[0027] For each time node, the deviation value between the comprehensive state characteristics and the benchmark operating parameters is calculated. According to the degree of influence of the data, an adjustment weight is assigned to each type of data to generate a weighted deviation result.

[0028] Preferably, the transmitting the adjustment signal to the control terminal comprises the following steps:

[0029] Use the standardized transformation method to map the weighted deviation results to the response range of the adjustment signal, and generate the corresponding adjustment signals through the feedback channel, including the following scenarios:

[0030] Current compensation channel: compensates for current deviation by adjusting the bypass resistor value and transmits the compensation signal through the current regulation module;

[0031] Temperature adjustment channel: adjust the temperature deviation by starting the heat dissipation device or heating device, and display the adjustment status on the temperature control interface;

[0032] Status prompt channel: prompts operation deviation through sound frequency change, and plays the frequency change prompt sound through the prompt device;

[0033] An operation record is generated based on the hierarchical evaluation mechanism according to the adjustment signal, including a primary evaluation that calculates deviations in real time and provides feedback for adjustment, an intermediate evaluation that periodically summarizes operation data and analyzes the effects, and an advanced evaluation that comprehensively analyzes long-term operation data and optimizes strategies.

[0034] Preferably, the present invention further includes a smart battery hybrid parallel access system for implementing the above-mentioned smart battery hybrid parallel access control method, wherein the system includes a hybrid architecture building module, a multi-source data processing module, and an operation deviation feedback module;

[0035] The hybrid architecture building module is used to divide the battery pack into a basic energy storage unit and a dynamic adjustment unit, and build a hybrid parallel access network based on a hierarchical interactive architecture;

[0036] The multi-source data processing module is used to collect unit voltage, loop current and temperature field distribution data through the multi-dimensional perception module, use the time series correlation processor to extract features and fuse weights of the multi-source data, input the fused data into the state decision module to adjust the access mode in real time, and execute the mode switching operation through the dual-path instruction parser;

[0037] The operation deviation feedback module is used for comparing and analyzing the real-time acquisition data with the benchmark parameters through a hierarchical evaluation mechanism, determining the operation deviation of the battery pack, transmitting the adjustment signal to the control terminal through the feedback channel and generating the operation record, and making the response speed of the adjustment signal positively correlated with the operation deviation.

[0038] Preferably, the hybrid architecture construction module comprises a function classification unit and a region allocation unit.

[0039] The function classification unit is used for setting the classification standard of the unit function, traversing each energy storage unit in the battery pack, evaluating the capacity characteristics and response ability thereof, and recording the function attributes of each unit, while identifying the units with charge and discharge state changes in the battery pack.

[0040] The region allocation unit is used for dynamically dividing the battery pack into a basic energy storage region and a dynamic adjustment region by using a region division strategy, assigning each function region to a different control node, generating a new global synchronization identifier according to the system operation frequency, and sending the global synchronization identifier to each control node, wherein each node performs an access parameter configuration operation according to the assigned function region and the current synchronization identifier, integrates the configuration results of all function regions into the main control network, and forms a complete hybrid parallel access architecture.

[0041] Compared with the prior art, the present application has the following advantages:

[0042] The intelligent battery hybrid parallel access system and control method have significant technical advantages. By dividing the battery pack into basic energy storage units and dynamic adjustment units, and constructing a hybrid parallel access network based on a hierarchical interaction architecture, the capacity characteristics and response ability of the units can be accurately classified, and the basic energy storage region and the dynamic adjustment region can be dynamically adjusted based on the region division strategy, so that the system can achieve efficient energy distribution and dynamic adjustment under different operating conditions, greatly improving the flexibility and adaptability of the system.

[0043] The multi-dimensional perception module acquires unit voltage, loop current and temperature field distribution data in real time, extracts features and fuses weights of multi-source data using a time sequence correlation processor, aligns different types of data in the time dimension and gives them differentiated weights through a clock synchronization module, and generates comprehensive state features, providing accurate and reliable data support for state decision making, effectively improving the accuracy and real-time performance of state decision making, and enabling timely adjustment of the access mode according to the battery state. The dual-channel instruction parser comprises a voltage instruction channel and a current instruction channel, which respectively use feature extraction and sequence decoding models, spatial correlation and time tracking models for semantic analysis, set confidence score to judge the effectiveness of the instructions, and in the abnormal signal processing scene, establish a voltage-current instruction mapping table for weighted fusion, ensuring the reliability and stability of the mode switching operation and reducing the occurrence of operation errors.

[0044] The hierarchical evaluation mechanism compares and analyzes the real-time collected data with the benchmark parameters, determines the operation deviation of the battery pack, transmits the adjustment signal to the control terminal through the current compensation channel, the temperature adjustment channel and the state prompting channel, the response speed of the adjustment signal is positively correlated with the operation deviation, the effective adjustment signal can be generated in time according to the deviation degree, the precise adjustment of the current, temperature and other parameters is realized, the operation record containing primary, intermediate and advanced evaluation is generated, comprehensive data support is provided for the optimized operation of the system, and the stability and safety of the system are effectively improved. The region division strategy recursively divides the sub-regions based on the core regulation region according to the running state of the battery pack, determines the configuration priority according to the correlation degree of the sub-region and the benchmark, and dynamically updates the division, so that the configuration of the control node is more reasonable, and the flexibility and adaptability of the overall access architecture are further improved. BRIEF DESCRIPTION OF DRAWINGS

[0045] Figure 1 The working principle diagram of the intelligent battery hybrid parallel access control method is described.

[0046] Figure 2 The flowchart of the region division strategy is described.

[0047] Figure 3 The fusion diagram of the time sequence correlation processor data is described.

[0048] Figure 4 The flowchart of the operation deviation determination is described. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0050] Please refer to Figure 1-Figure 4 The present application provides an intelligent battery hybrid parallel access system and control method, the specific implementation steps are as follows:

[0051] The battery pack is divided into a basic energy storage unit and a dynamic adjustment unit, and a hybrid parallel access network is constructed based on a hierarchical interactive architecture.

[0052] The single-unit voltage, loop current and temperature field distribution data are collected by a multi-dimensional perception module, the multi-source data are subjected to feature extraction and weight fusion by a time sequence correlation processor, the fused data are input into a state decision module to adjust the access mode in real time, and a dual-channel instruction parser is used to perform mode switching operation.

[0053] Through a hierarchical evaluation mechanism, the real-time collected data is compared and analyzed with the benchmark parameters to determine the operating deviation of the battery pack. The adjustment signal is transmitted to the control terminal through the feedback channel and an operation record is generated. The feedback channel includes a current compensation channel, a temperature adjustment channel and a status prompt channel. Among them, the response speed of the adjustment signal is positively correlated with the operating deviation.

[0054] Example 1:

[0055] In the process of building a hybrid parallel access network, it is necessary to set classification standards for unit functions. The classification standards mentioned here are determined based on the performance characteristics of each energy storage unit in the battery pack and its role in the system. Each energy storage unit in the battery pack is comprehensively traversed, and the capacity characteristics and responsiveness of each unit are evaluated one by one. The capacity characteristics mainly focus on the amount of electricity that the unit can store and the ability to maintain capacity during the charging and discharging process; the responsiveness focuses on the reaction speed and adaptability of the unit when facing changes in parameters such as current and voltage in the system. While evaluating, the functional attributes of each unit are recorded in detail. These attributes include information such as the unit type, specifications, and performance under different working conditions, so that the units can be classified and managed later.

[0056] During the cell traversal and evaluation process, it is necessary to identify cells in the battery pack whose charge and discharge states have changed. These changes in charge and discharge states can be caused by factors such as system load changes and the ongoing charging process. These changes in cell states can affect the performance of the entire battery pack and therefore require special attention. After identifying these cells, a regional partitioning strategy is used to dynamically divide the battery pack into a basic energy storage zone and a dynamic regulation zone.

[0057] The specific implementation steps of the zoning strategy are as follows: The monitoring module tracks the battery pack's operating status in real time. This monitoring includes key parameters such as charge and discharge power and temperature distribution. Changes in charge and discharge power reflect the battery pack's workload, while temperature distribution directly impacts battery safety and service life. Based on the real-time monitored operating status, a core regulation region is determined. This region typically represents the portion of the battery pack where operating status fluctuates most dramatically and has a significant impact on the entire system. Using the identified core region as a baseline, the set range is expanded outward, grouping the cells within this range. This range is determined based on the battery pack's specific structure, performance parameters, and actual operational requirements. Different application scenarios may require different expansion ranges.

[0058] The battery pack is recursively divided into sub-regions. This recursive division method makes the division more detailed and reasonable, better adapting to the complex operating conditions of the battery pack. The configuration priority of a sub-region is determined by its degree of correlation with the benchmark. The higher the correlation, the greater the influence of the sub-region on the core regulation region, and it needs to be given priority during configuration. Sub-regions with the same correlation level are coordinated through synchronization identifiers. The role of synchronization identifiers is to ensure that these sub-regions remain consistent and coordinated during the configuration process, avoiding confusion and conflicts. Furthermore, as the operating status of the battery pack continues to change, the sub-region division needs to be dynamically updated to ensure the rationality and effectiveness of the division, so that the system can always maintain optimal operating conditions.

[0059] After completing the area division, each functional area is assigned to a different control node. The control node is the part of the system responsible for controlling and managing each functional area. Assigning different functional areas to different control nodes allows for more precise control. A new global synchronization identifier is generated based on the system's operating frequency. The system operating frequency refers to the speed at which parameters change during system operation. The generated global synchronization identifier must reflect the system's current operating status. The global synchronization identifier is then sent to each control node. Each node performs access parameter configuration operations based on the assigned functional area and the current synchronization identifier. Access parameter configuration includes setting parameters such as voltage, current, and resistance. The configuration of these parameters directly affects the operating performance of the battery pack. The configuration results of all functional areas are integrated into the main control network, which is the core control component of the entire system. By integrating the configuration results of each functional area, a complete hybrid parallel access architecture is formed, enabling coordinated and stable operation of the entire battery pack.

[0060] Throughout the implementation process, each step must be strictly carried out in accordance with established standards and procedures to ensure seamless integration and accurate data recording and transmission. For example, when evaluating a cell's capacity characteristics and responsiveness, specialized testing equipment and methods are required to ensure accurate results. When dividing regions and allocating control nodes, the structural characteristics and operational requirements of the battery pack must be fully considered to ensure a more rational division and allocation. When generating and sending global synchronization identifiers and configuring access parameters, data transmission speed and reliability must be guaranteed to avoid delays and errors. This rigorous series of operations ultimately enables the construction of a hybrid parallel access network, ensuring the efficient and stable operation of smart batteries.

[0061] Example 2:

[0062] In the multi-source data processing link, the multi-dimensional perception module needs to perform differentiated data collection operations. For the unit voltage, the real-time collection method is adopted, that is, the voltage of each battery unit is continuously monitored to ensure the immediacy and continuity of the voltage data, so as to capture the instantaneous changes of the voltage. For the loop current, the periodic collection strategy is implemented, and the current is sampled at a pre-set time interval, which can be determined according to the required accuracy of the system for the current data and the actual running situation, such as collecting once every millisecond or every few seconds. While ensuring the effectiveness of the data, the over-collection leading to data redundancy is avoided. The temperature field distribution data is continuously collected, and the temperature information of each region is obtained through the temperature sensors distributed in different positions of the battery pack, so as to fully reflect the temperature distribution state and change trend of the battery pack.

[0063] After collecting various types of data, the clock synchronization module is used to align the different types of data in the time dimension. Due to the different collection frequencies and methods of voltage, current and temperature data, they may have deviations in the time axis. The clock synchronization module adjusts these data to the same time coordinate system through a unified time reference, so that different data have comparability and correlation in time. For example, the voltage data at a certain time, the current data of the corresponding period and the temperature data at that time are accurately matched, laying a foundation for subsequent comprehensive analysis.

[0064] Next, according to the stability and influence degree of the data, different weights are given to different data. The stability of the data refers to the fluctuation degree of the data in a period of time, and the stability is high when the fluctuation is small. The influence degree refers to the influence size of the change of the data on the overall operation state of the battery pack. For example, the sudden change of the voltage may directly affect the normal operation of the equipment, and the influence degree is relatively high, so a higher weight can be given. Although the temperature data is also important, the change is relatively slow, and the weight can be adjusted according to the actual situation. A composite processing structure is designed, which is composed of a storage unit and a calculation unit. The storage unit is used to process historical data, and the collected historical voltage, current and temperature data are stored and managed for subsequent query and analysis trend; the calculation unit focuses on analyzing real-time data and processing and calculating the current collected data.

[0065] Multi-source data is input into the storage unit and the computing unit, respectively. The storage unit classifies, organizes, and stores historical data to create a data archive. The computing unit extracts features from real-time data. Using specific algorithms and processing methods, it extracts key characteristic information from voltage, current, and temperature data that reflects the battery pack's operating status, such as voltage fluctuations, current trends, and temperature anomalies. The extracted feature information is then superimposed and fused based on pre-assigned weights, combining features from different data types according to the weighted ratio to generate a comprehensive status feature. This comprehensive status feature comprehensively and accurately reflects the overall operating status of the battery pack at the current moment.

[0066] The dual-path command parser consists of a voltage command channel and a current command channel, each responsible for different command parsing tasks. The voltage command channel uses a feature extraction and sequence decoding model for semantic parsing. The feature extraction module extracts voltage fluctuation characteristics and analyzes real-time voltage data to identify characteristics such as the pattern, amplitude, and frequency of voltage fluctuations. The sequence decoding module analyzes the temporal dependencies of voltage data, examining the temporal variations and interactions of voltage over time to understand the information contained in the voltage fluctuations. After processing, it outputs structured control instructions that contain specific adjustment solutions for voltage issues.

[0067] A recognition confidence scoring mechanism is also implemented to determine the validity of voltage commands. The recognition confidence score is determined based on the results of feature extraction and sequence decoding, reflecting the accuracy of the voltage command parsing. If the recognition confidence score falls below a set threshold, it indicates a significant error in the voltage command parsing and may not accurately reflect the actual situation. In this case, an abnormal signal is output. If the recognition confidence score is at least the set threshold, the parsing result is relatively reliable, and a switch command is output to execute the corresponding mode switch operation.

[0068] The current command channel performs semantic analysis based on spatial correlation and temporal tracking models. Continuously sampled loop current data is input, and the correlation analysis module captures the correlation between the spatial distribution and temporal variation of this current data. Spatial distribution refers to the distribution of current in different loops of the battery pack, while temporal variation describes the current fluctuation trend over time. By analyzing the correlation between these two, the causes and effects of current variations can be understood. After the analysis is complete, a current regulation code is output, which contains the regulation strategy for the current problem.

[0069] Similarly, an analysis confidence score is set to determine the validity of the current command. The analysis confidence score assesses the reliability of the current command analysis result. If the analysis confidence score is below the set threshold, it indicates uncertainty in the current command analysis and an abnormality signal is output. If the analysis confidence score is at least the set threshold, the analysis result is reliable and a switch command is output to achieve the mode switch.

[0070] Throughout the entire data processing and command parsing process, every step requires precise execution. During data collection, sensor accuracy and stability must be guaranteed to ensure the authenticity and reliability of the collected data. Clock synchronization must be accurate to avoid data errors caused by time deviations. Weights must be set appropriately, fully considering the actual impact of the data. Feature extraction and model analysis must utilize appropriate algorithms and methods to improve parsing accuracy. The confidence score threshold must be determined based on the system's actual needs and operational experience, balancing system reliability and responsiveness. The coordinated operation of these steps enables efficient processing of multi-source data and accurate parsing of commands, providing a basis for adjusting the battery pack's operating mode.

[0071] Example 3:

[0072] There are various application scenarios for the dual-path command parser to perform mode switching operations, each corresponding to different triggering conditions and processing flows. In scenario one, the system first receives unit voltage, loop current, and temperature field distribution data collected by the multi-dimensional perception module. This data directly reflects the system's operating status. Next, the loop current data is processed to generate continuous sampling data. This continuous sampling of current data allows for more detailed observation of current trends and anomalies. Based on this continuous sampling data, the current command channel is triggered to begin operation. Based on its internal spatial correlation and temporal tracking model, the current command channel performs semantic analysis on the current data, capturing correlations between the spatial distribution and temporal variations of the current, and then outputs a current regulation code. When the analysis confidence score corresponding to the current regulation code is at least a set threshold, a switching command is generated. The system executes the corresponding mode switching operation based on this switching command to adapt to the impact of current fluctuations and ensure the normal operation of the battery pack.

[0073] The trigger condition for scenario two is receiving real-time voltage fluctuation data. The multi-dimensional perception module collects unit voltage data in real time and transmits the voltage fluctuation information to the system in real time. After receiving this real-time voltage fluctuation data, the feature extraction module of the voltage command channel extracts the voltage fluctuation characteristics and analyzes characteristic information such as the amplitude, frequency, and pattern of the voltage fluctuation. The sequence decoding module then analyzes the temporal dependencies of the voltage data to understand the patterns of voltage changes and their mutual influence over time. Through the collaborative work of these two modules, a structured adjustment command is output. The adjustment command is also evaluated for recognition confidence. If the score is at least a set threshold, the command is valid, triggering the voltage command channel to output a switching command. The system then executes a mode switch based on this switching command to address the issues caused by voltage fluctuations and ensure that the battery pack voltage remains stable within a reasonable range.

[0074] Scenario three is a processing flow initiated when abnormal signals are detected in the output of the voltage command channel and the current command channel. When the system receives abnormal signals from the two channels, it first determines whether the number of abnormal signals exceeds the warning threshold. The warning threshold is a critical value pre-set based on the actual operation of the system and historical data, which is used to determine the severity of the abnormal situation. If the number of abnormal signals exceeds the threshold, it indicates that the system may have a more serious fault or abnormal situation, which requires more in-depth processing. At this time, the system will establish a voltage-current command mapping table, which is used to record the correspondence and correlation between the voltage command and the current command.

[0075] Next, the confidence scores of the voltage and current channels are weighted and fused. The weights used in weighted fusion are pre-set based on the importance of voltage and current in the system and the degree of mutual influence between them. Through weighted fusion, a comprehensive confidence score is obtained, which reflects the degree of match between the dual-path instructions and the regulation strategy. The comprehensive confidence score is then compared with the overall threshold, which is also pre-set based on system requirements. If the comprehensive confidence score exceeds the overall threshold, it means that despite the presence of abnormal signals, the overall effectiveness of the dual-path instructions is still high. The instructions are judged to be valid and the mode switch is triggered, allowing the system to maintain normal operation as much as possible under abnormal conditions. If the comprehensive confidence score does not exceed the overall threshold, it indicates that the effectiveness of the instructions is insufficient. The system sends a warning signal to the control terminal, reminding relevant personnel to pay attention to the operating status of the battery pack so that timely inspection and maintenance can be carried out.

[0076] Throughout the implementation process, all aspects of each scenario require close coordination. Data collection and transmission must be real-time and accurate to ensure the system can promptly obtain true operating status information; command channel analysis and processing must be efficient and precise to ensure that the output commands accurately reflect the actual needs of the system; confidence scores must be set and compared reasonably to ensure that correct decisions can be made in different situations; and abnormal situations must be handled with caution, ensuring both system stability and reliability while avoiding misjudgments and missed detections. For example, in scenario three, the setting of warning thresholds and overall thresholds must fully consider the system's actual operating environment and possible abnormal situations, and the distribution of weights must reflect the relative importance of voltage and current in the system. Through strict control of these details, the dual-channel command parser can ensure that mode switching operations are correctly executed in various scenarios, keeping the battery pack in optimal operating condition and improving system stability and reliability.

[0077] Example 4:

[0078] When determining the battery pack's operating deviation, the first step is to establish baseline operating parameters as control targets. These parameters include a baseline voltage range, a baseline current threshold, and a baseline temperature range. These parameters are set based on the battery pack's design specifications, normal operating conditions, and actual application requirements. For example, the baseline voltage range for a particular battery pack model might be set to 3.6V to 4.2V, the baseline current threshold might be set to 10A and 5A, depending on the charge and discharge mode, and the baseline temperature range might be 20°C to 40°C, ensuring that the battery operates within a safe and effective range.

[0079] After establishing the baseline operating parameters, they need to be aligned with the battery pack's real-time operating data in the time dimension. Since the baseline parameters are fixed reference ranges, while the real-time data changes over time, time alignment allows the real-time data at each moment to be accurately compared with the corresponding baseline parameters. For example, at a certain moment t1, the real-time collected cell voltage is 3.8V, the loop current is 8A, and the temperature is 25°C. At this time, these data are aligned with the baseline voltage range of 3.6V-4.2V, the baseline current threshold (such as no more than 10A when charging), and the baseline temperature range of 20°C-40°C in time for subsequent analysis.

[0080] For each time point, the deviation between the comprehensive state characteristics and the baseline operating parameters is calculated. The comprehensive state characteristics are derived by extracting and weighting data such as voltage, current, and temperature through the multi-source data processing module, and can fully reflect the operating status of the battery pack. Taking voltage as an example, if the voltage characteristic value in the comprehensive state characteristics at a certain moment is 3.5V and the baseline voltage lower limit is 3.6V, the voltage deviation value is -0.1V; if the current characteristic value is 12A and the baseline current threshold is 10A, the current deviation value is +2A.

[0081] An adjustment weight is assigned to each data type based on its impact. Different data types have different impacts on battery pack operation. For example, voltage deviation can directly affect device performance, so its adjustment weight may be set higher. Temperature deviation, while affecting battery life, has a relatively slow response time, so its weight can be appropriately reduced. Assuming the adjustment weights for voltage, current, and temperature are set to 0.5, 0.3, and 0.2, respectively, a weighted calculation is performed to convert the deviation values ​​of each data type into a weighted deviation. For example, if the voltage deviation is -0.1V, the current deviation is +2A, and the temperature deviation is +5°C (assuming the real-time temperature is 45°C), the weighted deviation is (-0.1 × 0.5) + (2 × 0.3) + (5 × 0.2) = (-0.05) + 0.6 + 1 = 1.55.

[0082] When transmitting the control signal to the control terminal, a standardized conversion method is used to map the weighted deviation result to the control signal's response range. This method applies a linear or nonlinear conversion to the weighted deviation result based on the system's set control signal range to ensure it meets signal transmission requirements. For example, if the control signal's response range is set to 0-10V and the weighted deviation result is 1.55, the corresponding control signal voltage value, such as 3V, is calculated using the conversion formula.

[0083] Corresponding adjustment signals are generated through the feedback channel. Specifically, in the following scenario: In the current compensation channel, when the current deviation is +2A, the bypass resistor value needs to be adjusted to compensate for the current deviation. Assuming the bypass resistor is 10Ω during normal operation, the current deviation determines that the resistor needs to be adjusted to 8Ω to divert excess current and return the loop current to the baseline threshold. This compensation signal is then transmitted through the current regulation module to the relevant actuators.

[0084] In the temperature control channel, if the real-time temperature reaches 45°C, exceeding the upper limit of the reference temperature range of 40°C, a cooling device, such as a fan or air conditioner, is activated to cool the battery pack. Simultaneously, the temperature control interface displays the adjustment status, for example, "Cooling device activated, current temperature 45°C, target temperature 40°C," allowing operators to monitor the temperature adjustment status in real time.

[0085] In the status prompt channel, when there is an operating deviation, a sound frequency change is used to indicate it. For example, when the voltage deviation is -0.1V, the tone frequency might be set to 500Hz. As the deviation increases, the frequency increases accordingly, reaching 1000Hz. The operator can intuitively understand the severity of the deviation through the changing tone frequency. The prompt device plays a tone with changing frequency, providing real-time notification of operating deviations.

[0086] Based on the adjustment signals, an operation record is generated using a hierarchical evaluation mechanism. The primary evaluation involves real-time deviation calculation and feedback adjustment, for example, recording the deviation value of each data point and the corresponding adjustment measures once a second. The intermediate evaluation periodically summarizes operation data and analyzes the results, such as daily summary of daily deviation data and analysis of the effectiveness of adjustment measures. The advanced evaluation comprehensively analyzes long-term operation data and optimizes strategies, such as conducting a comprehensive analysis of three months of operation data every quarter and adjusting benchmark parameters, adjustment weights, and other strategies based on actual conditions to improve system efficiency and stability.

[0087] Throughout the implementation process, benchmark parameters must be set based on reference to the battery pack's technical documentation and actual operating experience to ensure their rationality. Time alignment must be precise to avoid deviation calculation errors caused by timing errors. Adjustment weights must be assigned by comprehensively considering the impact of various data points and determined through multiple tests and optimizations. Standardized conversion methods must ensure the accuracy and reliability of adjustment signals. Adjustment operations in the feedback channel must be timely and effective to ensure timely adjustments to the battery pack's operating status. A tiered evaluation mechanism must be strictly implemented, with evaluations conducted over different periods to comprehensively understand the battery pack's operating status and provide a basis for system optimization. For example, when setting the benchmark temperature range, factors such as battery type and operating environment must be considered. For batteries used in high-temperature environments, the upper limit of the benchmark temperature range can be appropriately increased. When assigning adjustment weights, if the system requires high voltage stability, the voltage adjustment weight can be set to 0.6, the current weight to 0.2, and the temperature weight to 0.2 to emphasize the importance of voltage regulation. Through meticulous attention to each link, accurate determination and effective adjustment of battery pack operating deviations are achieved, ensuring safe and stable operation of the battery pack.

[0088] Example 5:

[0089] The intelligent battery hybrid parallel access system is used to implement the above-mentioned control method. The construction of this system revolves around the hybrid architecture construction module, the multi-source data processing module, and the operation deviation feedback module. Taking the battery pack system of a certain energy storage power station as an example, the hybrid architecture construction module needs to divide the battery pack into basic energy storage units and dynamic adjustment units. Assuming that the battery pack of the power station consists of 100 energy storage units, the module first sets the classification standard of the unit function, such as using a capacity retention rate of more than 80% and a response time of less than 50ms as the screening conditions for basic energy storage units. When traversing each unit, the charge and discharge capacity characteristics are evaluated by the capacity testing equipment, and the response time detection device is used to measure its response ability to current changes.

[0090] During the evaluation process, cells with changing charge and discharge states are identified. For example, if a cell switches charge and discharge states more than three times within 10 minutes, it is marked as a state-changing cell. When adopting a regional division strategy, the charge and discharge power and temperature distribution of the battery pack are tracked in real time through temperature sensors and power monitoring devices. If it is found that the temperature of 5 cells in a certain area is continuously 5°C higher than that of other areas within 30 minutes and the charge and discharge power fluctuation exceeds the average level by 20%, then this area is determined as the core adjustment area. Based on this area, the range of 2 cells is expanded outward to form a sub-area. After recursive division, the battery pack is divided into 5 sub-areas. The configuration priority is set according to the distance between the sub-area and the core area. The closer the distance, the higher the priority. Sub-areas with the same priority are coordinated through synchronization identification.

[0091] The regional allocation unit assigns the basic energy storage area and the dynamic adjustment area to different control nodes. For example, node A controls 30 units in the basic energy storage area, and node B controls 70 units in the dynamic adjustment area. A global synchronization identifier, such as "SYNC_20250630_100Hz," is generated based on the system's 100Hz operating frequency. After sending it to each node, node A configures access parameters based on the assigned functional area and synchronization identifier. For example, it sets the voltage upper limit of the basic energy storage area units to 4.0V and the current compensation threshold of the dynamic adjustment area units to ±15A. Node B then integrates the configuration results into the main control network, forming a hybrid parallel access architecture.

[0092] In the multi-source data processing module, the multi-dimensional perception module collects unit voltages in real time. For example, high-precision voltage sensors collect voltage data for each unit at a frequency of 1000Hz. Loop current is collected at a 100Hz cycle, sampled every 10ms using current transformers. Temperature distribution is continuously collected using distributed temperature sensors, recording the temperature at each point once per second. The clock synchronization module uses the main control network's clock as a reference to align voltage, current, and temperature data to the same timeline. For example, voltage data at t=10,000s, current periodic sampling data from t=10,000s to 10,010s, and temperature data at t=10,000s are marked as data from the same time node.

[0093] Weights are assigned based on data characteristics. Voltage data has a direct impact on device operation, so its weight is set to 0.4. Current data affects charge and discharge efficiency, so its weight is set to 0.3. Temperature data affects battery life, so its weight is set to 0.3. The storage unit of the composite processing structure stores historical data from the past 24 hours, while the computing unit analyzes current data in real time, extracting features such as voltage fluctuation amplitude, current change rate, and temperature anomalies. These features are then superimposed and integrated to generate comprehensive state features. The voltage command channel of the dual-path command parser extracts voltage fluctuation features. For example, if the voltage of a unit drops from 3.8V to 3.5V within 200ms, with a fluctuation amplitude of 0.3V, the sequence decoding module analyzes the timing dependency of this voltage drop to determine whether it is caused by a sudden increase in load. The output adjustment command is to start the backup unit in parallel. If the recognition confidence score reaches 85% (the set threshold is 70%), a switching command is output.

[0094] The current command channel receives continuous sampling data of the loop current. For example, if the current in a certain loop increases from 80A to 120A within 500ms, the correlation analysis module will detect a simultaneous 3°C temperature rise in the sub-region where the loop is located, and determine that it is overloaded. The output current adjustment code is to increase the bypass resistance of the loop to 10Ω. The analysis confidence score is 90% (threshold 75%), and a switching command is output.

[0095] The operating deviation feedback module establishes baseline operating parameters, such as a baseline voltage range of 3.6V-4.2V, a baseline current threshold of 100A for charging and 80A for discharging, and a baseline temperature range of 25°C-35°C. At a certain moment, the comprehensive state characteristics indicate a voltage of 3.5V, a current of 110A, and a temperature of 38°C. The calculated deviations from the baseline parameters are -0.1V, +30A, and +3°C, respectively. Using weights of 0.4, 0.3, and 0.3, the weighted deviation is (-0.1 × 0.4) + (30 × 0.3) + (3 × 0.3) = -0.04 + 9 + 0.9 = 9.86.

[0096] Through standardized conversion, 9.86 is mapped to the 0-10V regulation signal range, resulting in a signal value of approximately 9.9V. The current compensation channel adjusts the bypass resistor from 5Ω to 8Ω, shunting 20A to reduce the current to 90A. The temperature adjustment channel activates the cooling fan and displays the temperature control interface as "Cooling, current 38°C, target 35°C." The status prompt channel generates an 800Hz beep (normal frequency is 500Hz). A hierarchical assessment mechanism generates operational records. The primary assessment records deviations and adjustment measures every second. For example, if the deviation is 9.86 at t=10.0s, the resistor is adjusted to 8Ω. The intermediate assessment summarizes the results daily. For example, if there are 10 current overloads that day, the average adjustment time is 200ms. The advanced assessment analyzes data quarterly and adjusts the baseline current threshold to 85A based on three months of data to adapt to load changes.

[0097] When the various modules of the system work together, the functional classification unit of the hybrid architecture building module must accurately record the attributes of each unit, such as unit 1 with a capacity of 100Ah and a response time of 40ms, marked as a basic energy storage unit. The regional allocation unit ensures that the control nodes configure parameters according to the synchronization identifier to avoid parameter conflicts between different regions. The sensor accuracy of the multi-source data processing module must be regularly calibrated. For example, the voltage sensor error must not exceed ±0.01V, and the time deviation of the clock synchronization module must not exceed ±1ms to ensure accurate data alignment. The baseline parameters of the operation deviation feedback module must be regularly updated according to the degree of battery aging, such as adjusting the baseline voltage lower limit to 3.5V after one year of use. The adjustment device of the feedback channel must ensure a fast response speed, such as the cooling fan starting within 500ms after receiving the signal. Through this specific implementation method, the system realizes intelligent management of the battery pack and ensures its stable operation in the hybrid parallel access mode.

[0098] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0099] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling hybrid parallel connection of intelligent batteries, characterized in that: The steps include: The battery pack is divided into basic energy storage units and dynamic adjustment units, and a hybrid parallel access network is constructed based on a hierarchical interactive architecture; The multi-dimensional sensing module collects unit voltage, loop current and temperature field distribution data, and uses a time-series correlation processor to extract features and fuse weights from multi-source data. The fused data is input into the state decision module to adjust the access mode in real time, and the dual-path instruction parser executes the mode switching operation. A hierarchical assessment mechanism compares and analyzes real-time collected data with benchmark parameters to determine the operating deviation of the battery pack. Adjustment signals are transmitted to the control terminal through feedback channels, including current compensation, temperature adjustment, and status prompt channels, and an operating record is generated. The response speed of the adjustment signal is positively correlated with the operating deviation. The layered interaction architecture constructs a hybrid parallel access network, including the following steps: Set the classification criteria for unit functions, traverse each energy storage unit in the battery pack, evaluate its capacity characteristics and response capabilities, and record the functional attributes of each unit. At the same time, identify the units in the battery pack with changing charge and discharge states, and use a regional partitioning strategy to dynamically divide the battery pack into a basic energy storage area and a dynamic adjustment area; Each functional area is assigned to a different control node. A new global synchronization identifier is generated according to the system operating frequency and sent to each control node. Each node performs access parameter configuration operations based on the assigned functional area and the current synchronization identifier, and integrates the configuration results of all functional areas into the main control network to form a complete hybrid parallel access architecture. The dual-path instruction parser includes a voltage instruction channel and a current instruction channel. The mode switching operation is performed by the dual-path instruction parser, which includes the following steps: The voltage command channel uses a feature extraction and sequence decoding model for semantic parsing. The feature extraction module extracts voltage fluctuation characteristics, and the sequence decoding module analyzes timing dependencies, outputs structured adjustment commands, and sets a recognition confidence score to determine whether the voltage command is valid. If the recognition confidence score is lower than the set threshold, an abnormal signal is output. If the recognition confidence score is not lower than the set threshold, a switching command is output. The current command channel performs semantic analysis based on the spatial correlation and time tracking model, inputs continuous sampling data of the loop current, captures the correlation between spatial distribution and time change through the correlation analysis module, outputs the current adjustment code, and sets the analysis confidence score for judging the effectiveness of the current command. If the analysis confidence score is lower than the set threshold, an abnormal signal is output. If the analysis confidence score is not lower than the set threshold, a switching command is output.

2. The intelligent battery hybrid parallel access control method according to claim 1, characterized in that: The area division strategy includes the following steps: The monitoring module tracks the operating status of the battery pack in real time, including charge and discharge power and temperature distribution. The core adjustment area is determined according to the operating status. Taking the core area as the benchmark, the unit combination within the set range is expanded outward, and the battery pack is recursively divided into sub-areas. The degree of correlation between the sub-area and the benchmark is used as the basis for configuration priority. Sub-areas with the same degree of correlation are coordinated through synchronization identification, and the sub-area division is dynamically updated according to changes in the operating status.

3. The intelligent battery hybrid parallel access control method according to claim 1, characterized in that: The time series association processor performs feature extraction and weight fusion on multi-source data, including the following steps: Real-time collection of unit voltage data, periodic collection of loop current data, and continuous collection of temperature field distribution data; Different types of data are aligned in the time dimension through the clock synchronization module, and differentiated weights are assigned according to the stability and impact of the data. A storage unit for processing historical data and a computing unit for analyzing real-time data are designed to form a composite processing structure. Multi-source data are input into the storage unit and the computing unit respectively, and the feature information of the multi-source data is extracted. The information is superimposed and fused according to the weight to generate comprehensive state features.

4. The intelligent battery hybrid parallel access control method according to claim 1, characterized in that: The parser performs mode switching operations, including the following scenarios: Scenario 1: Receive the collected unit voltage, loop current and temperature field distribution data, generate continuous sampling data of the loop current, trigger the current command channel to output the current adjustment code, and execute the mode switching operation according to the switching command; Scenario 2: Receive real-time voltage fluctuation data, extract voltage fluctuation characteristics, trigger the voltage command channel to output structured adjustment instructions, and execute mode switching operations according to the switching instructions; Scenario 3: Sense abnormal signals output by the voltage command channel and the current command channel. If the number of abnormal signals exceeds the warning threshold, establish a voltage-current command mapping table, perform weighted fusion on the confidence scores of the voltage channel and the current channel, output the degree of matching between the dual-channel command and the regulation strategy, and set the comprehensive confidence. If the comprehensive confidence exceeds the total threshold, the command is determined to be valid and triggers mode switching. Otherwise, send a warning signal to the control terminal.

5. The intelligent battery hybrid parallel access control method according to claim 3, characterized in that: Determining the operating deviation of the battery pack includes the following steps: Establish baseline operating parameters as control targets, including a baseline voltage range, a baseline current threshold, and a baseline temperature range, and align the baseline operating parameters with the battery pack's real-time operating data in the time dimension; For each time node, the deviation value between the comprehensive state characteristics and the benchmark operating parameters is calculated. According to the degree of influence of the data, an adjustment weight is assigned to each type of data to generate a weighted deviation result.

6. The intelligent battery hybrid parallel access control method according to claim 5, characterized in that: The step of transmitting the adjustment signal to the control terminal comprises the following steps: Use the standardized transformation method to map the weighted deviation results to the response range of the adjustment signal, and generate the corresponding adjustment signals through the feedback channel, including the following scenarios: Current compensation channel: compensates for current deviation by adjusting the bypass resistor value and transmits the compensation signal through the current regulation module; Temperature adjustment channel: adjust the temperature deviation by starting the heat dissipation device or heating device, and display the adjustment status on the temperature control interface; Status prompt channel: prompts operation deviation through sound frequency change, and plays the frequency change prompt sound through the prompt device; An operation record is generated based on the hierarchical evaluation mechanism according to the adjustment signal, including a primary evaluation that calculates deviations in real time and provides feedback for adjustment, an intermediate evaluation that periodically summarizes operation data and analyzes the effects, and an advanced evaluation that comprehensively analyzes long-term operation data and optimizes strategies.

7. A smart battery hybrid parallel access system, used to implement the smart battery hybrid parallel access control method according to any one of claims 1 to 6, characterized in that: It includes a hybrid architecture building module, a multi-source data processing module, and an operation deviation feedback module; The hybrid architecture building module is used to divide the battery pack into a basic energy storage unit and a dynamic adjustment unit, and build a hybrid parallel access network based on a hierarchical interactive architecture; The multi-source data processing module is used to collect unit voltage, loop current and temperature field distribution data through the multi-dimensional perception module, use the time series correlation processor to extract features and fuse weights of the multi-source data, input the fused data into the state decision module to adjust the access mode in real time, and execute the mode switching operation through the dual-path instruction parser; The operation deviation feedback module is used to compare and analyze the real-time collected data with the benchmark parameters through a hierarchical evaluation mechanism, determine the operation deviation of the battery pack, transmit the adjustment signal to the control terminal through the feedback channel and generate an operation record, and make the response speed of the adjustment signal positively correlated with the operation deviation.

8. The intelligent battery hybrid parallel access system according to claim 7, characterized in that: The hybrid architecture building module includes a functional classification unit and a regional allocation unit; The functional classification unit is used to set the classification standard of the unit function, traverse each energy storage unit in the battery pack, evaluate its capacity characteristics and response capabilities, and record the functional attributes of each unit. At the same time, it identifies the units in the battery pack whose charge and discharge states have changed; The regional allocation unit is used to dynamically divide the battery pack into a basic energy storage area and a dynamic adjustment area using a regional division strategy. Each functional area is assigned to a different control node, a new global synchronization identifier is generated according to the system operating frequency, and the global synchronization identifier is sent to each control node. Each node performs access parameter configuration operations based on the assigned functional area and the current synchronization identifier, and integrates the configuration results of all functional areas into the main control network to form a complete hybrid parallel access architecture.

Images