A bidirectional charging and discharging system and method for a battery pack based on energy mutual feeding

Abstract

The present application relates to the technical field of battery testing, and particularly relates to a bidirectional charging and discharging system and method for storage battery group based on energy mutual feedback, which comprises the following steps: connecting multiple groups of storage battery groups to be tested to a charging and discharging test field which is configured with a bidirectional conversion module and an energy buffer unit, and extracting task characteristic parameters; obtaining an initial operation strategy in combination with a preset historical working condition knowledge base; calling a test process file bound to each storage battery group to be tested, analyzing power parameters and time parameters of each working step in the test process file, and generating a test timing planning scheme according to an operation result; driving the bidirectional conversion module to work, monitoring a power difference value of energy transmission in real time, and dynamically balancing electric energy from the energy buffer unit according to a power difference value state; collecting energy operation data in real time, obtaining a mutual feedback rate through a preset mutual feedback rate calculation formula, and generating a test resource scheduling instruction in combination with a preset scheduling rule base. The present application greatly improves the energy utilization rate in the field through energy mutual feedback matching and scheduling of multiple battery groups.

Description

Technical Field

[0001] This invention relates to the field of battery testing technology, and in particular to a bidirectional charging and discharging system and method for a battery pack based on energy mutual feedback. Background Technology

[0002] Currently, various battery packs, including lead-acid batteries, lithium batteries, and nickel-metal hydride batteries, generally adopt a single-cell series charge-discharge mode in cycle charge-discharge tests, capacity verification, and formation processes. Traditional solutions rely on resistive loads to dissipate discharge energy, resulting in significant energy waste and requiring additional cooling devices, leading to high operating costs.

[0003] While existing energy feedback technologies can transmit discharged electrical energy to the grid and alleviate energy consumption problems, they still have significant shortcomings. Most devices only support independent operation of a single battery group, and testing multiple battery groups can only be carried out sequentially, resulting in long testing cycles and low operational efficiency. Furthermore, the equipment cannot automatically determine the battery's operating status, and switching between charging and discharging roles relies on manual operation, making unattended continuous operation difficult.

[0004] Furthermore, most existing energy feedback schemes rely on the power grid as an energy supplement or recovery carrier, making them unusable in off-grid scenarios such as outdoor mobile detection, remote base station maintenance, and on-site emergency repairs. Some schemes add extra energy storage devices to balance energy differences between battery banks, leading to complex system architecture and increased hardware costs. There is still significant room for improvement in existing technologies to meet requirements such as direct energy feedback between multiple battery banks, adaptive adjustment of energy differences, and stable operation under off-grid conditions. Summary of the Invention

[0005] This invention solves the problem of high energy loss in traditional battery charging and discharging tests by using energy feedback matching and scheduling among multiple battery packs, thus significantly improving the energy utilization rate in the field.

[0006] The technical solution proposed in this invention is: a bidirectional charging and discharging method for a battery pack based on energy mutual feedback, the method comprising: Multiple sets of battery packs under test are connected to a charge-discharge test field equipped with a bidirectional conversion module and an energy buffer unit to extract task characteristic parameters. Based on task feature parameters and a pre-set historical operating condition knowledge base, the optimal scheduling strategy is matched and obtained. The optimal scheduling strategy is then corrected based on the initial operating state of the test field to obtain the initial operating strategy. Retrieve the test process files bound to each battery pack under test, parse the power and time parameters of each step in the test process files, combine the initial running strategy, predict the power change trend corresponding to each step, and generate a test timing plan through a greedy algorithm. Based on the initial operation strategy and test timing plan, the bidirectional conversion module is driven to work, and the power difference of energy transmission is monitored in real time. According to the power difference status, the energy buffer unit performs dynamic power balancing. During charging and discharging operation, energy operation data is collected in real time. The energy operation data is calculated using a preset mutual feedback rate calculation formula to obtain the mutual feedback rate. Combined with a preset scheduling rule base, test resource scheduling instructions are generated to adjust the test task operation status.

[0007] Preferably, the process for obtaining the task feature parameters is as follows: Establish a full-domain data acquisition link to synchronously collect multi-dimensional raw operating condition data from multiple sets of batteries under test, bidirectional conversion modules, and test stations that have been connected to the charge and discharge test site. Data preprocessing is performed on the multi-dimensional raw working condition data. The cleaned multi-dimensional raw working condition data is then aligned in dimensions, standardized in units, and fused in features to construct a structured feature dataset and obtain task feature parameters.

[0008] Preferably, the process of obtaining the optimal scheduling strategy is as follows: The task feature parameters are used as a multi-dimensional retrieval index. The similarity matching algorithm is called to traverse the preset historical working condition knowledge base, compare the current working condition with the historical working condition, calculate the feature matching score in combination with the working condition coupling degree judgment standard, compare the feature matching score with the preset matching threshold, and filter out the set of the same source historical operation schemes that are higher than the matching threshold. Within the set of historical operation schemes from the same source, the scheme with the highest feature matching score is selected as the optimal scheduling strategy.

[0009] Preferably, the process of obtaining the initial running strategy is as follows: The environmental parameters of the entire charging and discharging test site are collected as constraints for the working condition correction. The optimal scheduling strategy is used as the baseline scheme. The baseline scheme is adapted and adjusted by an adaptive correction model in combination with the current working condition characteristics. The feasibility and adaptability of the adjusted scheme are verified by checking the power outage due to operating conditions, and an initial operating strategy adapted to the characteristics of the current operating conditions is output.

[0010] Preferably, the specific details of the test timing planning scheme are as follows: The standardized test process files bound to each battery pack under test are retrieved, and the core timing parameters corresponding to each test step are separated through semantic parsing and parameter decomposition. Using the initial operating strategy as a constraint boundary and combining the battery charging and discharging characteristics deduction algorithm, the power change trend under different work step switching nodes is predicted; The power change trend is input into a greedy algorithm, and the timing solution is performed with the goal of maximizing the energy mutual feedback rate in the field. Based on the solution results, a test timing planning scheme is generated.

[0011] Preferably, the dynamic power balancing process is as follows: Based on the initial operation strategy and test timing plan, the bidirectional conversion modules of each group are driven in a coordinated manner to carry out closed-loop energy mutual feed transmission between multiple battery groups. Collect real-time power data of the energy transmission link, calculate the power deviation on both sides of the charge and discharge, and determine the supply and demand status of the system; The hierarchical balancing control logic is invoked, and the corresponding adjustment level is matched according to the power deviation to control the energy buffer unit to perform corresponding operations, thereby achieving dynamic power balancing.

[0012] Preferably, the process of obtaining the test resource scheduling instruction is as follows: Under the operating condition of dynamic power balance, the energy operation data of the charge and discharge test field are continuously collected, and the mutual feedback rate is calculated according to the mutual feedback rate calculation formula. The feedback rate is integrated into the hierarchical scheduling rule base. The operating condition level is divided based on the deviation of the indicator and the duration of the abnormality. Test resource scheduling instructions are generated according to the operating condition level.

[0013] Preferably, the method further includes system iterative optimization, the specific process of which is as follows: After all the battery packs under test have completed the full charge and discharge cycle, the operating condition information, the feedback rate throughout the cycle, and the iterative operation strategy adjusted by instructions are collected to form the complete operation dataset for this cycle. After the complete dataset of this round of operation is structured, labeled with features and validated, it is written into the historical working condition knowledge base according to the knowledge base storage specifications to complete the iterative update of the historical working condition knowledge base; Once the knowledge base is updated, you will be automatically redirected to the next test task launch phase.

[0014] The present invention also provides a bidirectional charging and discharging system for a battery pack based on energy mutual feedback, the system being used to execute the aforementioned bidirectional charging and discharging method for a battery pack based on energy mutual feedback.

[0015] The present invention also provides a computer-readable storage medium storing a computer program that is executed by a processor to implement the aforementioned bidirectional charging and discharging method for a battery pack based on energy mutual feedback.

[0016] The beneficial effects of this invention are: 1. This invention establishes a multi-battery pack energy feedback matching and scheduling mechanism, matches the optimal historical scheduling scheme based on the similarity of operating conditions, and achieves precise peak-shifting and overlapping matching of charging and discharging steps using a greedy timing arrangement strategy. This changes the traditional operating mode of battery testing, where each battery pack operates independently, all discharge energy is dissipated, or the system relies on grid feedback. It enables direct energy exchange and utilization between multiple battery packs within the testing field. This effectively reduces the energy intervention ratio of the grid and backup batteries, significantly reduces energy loss during testing, and shortens the overall working time for multiple batches of battery testing, significantly improving the overall efficiency of battery charging and discharging testing, formation, and capacity verification.

[0017] 2. This invention employs a multi-level power deviation tiered equalization control logic. Based on the real-time power deviation magnitude within the testing area, it utilizes a layered control approach, including module duty cycle fine-tuning, buffer unit energy replenishment, and low-priority task pause. This enables differentiated and precise adjustment under small, medium, and large power imbalance conditions, solving the problems of unquantifiable power fluctuation control, energy supply and demand imbalance, and poor system stability in traditional testing systems. It effectively suppresses bus power fluctuations, prevents frequent over-threshold operation of batteries, ensures that each battery group and bidirectional converter module remain within a safe and stable operating range, and improves the overall reliability and adaptability of the testing equipment.

[0018] 3. This invention effectively filters, archives, and iteratively reuses test data, eliminating abnormal and invalid data that pollutes the knowledge base. Simultaneously, it directly reuses the optimal operating strategy from the current round and adaptively fine-tunes it as the initial baseline strategy for the next round of testing. This breaks the limitations of traditional testing schemes that operate with fixed parameters and lack self-learning optimization capabilities, achieving a closed-loop operating mode of continuous accumulation of test data and continuous adaptation and optimization of scheduling strategies. This enables the system to adapt to long-term testing needs of different battery types, capacities, and architectural conditions, continuously improving energy feedback rate and scheduling accuracy, and meeting the requirements of long-term, unattended, multi-batch automated testing operations. Attached Figure Description

[0019] Figure 1 A flowchart of a bidirectional charging and discharging system and method for a battery pack based on energy mutual feedback; Figure 2 A flowchart of a multi-battery pack energy mutual feedback matching and scheduling process for a bidirectional charging and discharging system and method based on energy mutual feedback. Figure 3 A flowchart illustrating the hierarchical dynamic balance control of energy in a bidirectional charging and discharging system and method for a battery pack based on energy mutual feedback. Figure 4 This is an iterative closed-loop flowchart of the operating condition knowledge base for a bidirectional charging and discharging system and method for a battery pack based on energy mutual feedback. Detailed Implementation

[0020] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.

[0021] It is understood that the term "a" should be understood as "at least one" or "one or more," that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0022] like Figure 1 As shown, a bidirectional charging and discharging method for a battery pack based on energy mutual feedback is provided, the method comprising: Multiple sets of battery packs under test are connected to a charge-discharge test field equipped with a bidirectional converter module and an energy buffer unit. Task feature parameters are extracted. Based on the task feature parameters and a pre-set historical operating condition knowledge base, an optimal scheduling strategy is obtained. The optimal scheduling strategy is then modified based on the initial operating state of the test field to obtain an initial operating strategy. The test process files bound to each battery pack under test are retrieved, and the power and time parameters of each step in the test process files are analyzed. Based on the initial operating strategy, the power change trend corresponding to each step is predicted, and a test timing plan is generated using a greedy algorithm. Based on the initial operating strategy and the test timing plan, the bidirectional converter module is driven to work, and the power difference in energy transmission is monitored in real time. According to the power difference status, the energy buffer unit performs dynamic energy balancing. During the charge-discharge operation, energy operation data is collected in real time. The energy operation data is calculated using a pre-set mutual feedback rate calculation formula to obtain the mutual feedback rate. Based on the pre-set scheduling rule base, test resource scheduling instructions are generated to adjust the test task operation status.

[0023] The charge / discharge test site in this embodiment includes three hardware architectures: Architecture 1 is an AC bus coupling type, which uses AC / DC bidirectional modules with contactor / relay switching circuits to connect to the power grid; Architecture 2 is a DC bus coupling type, which uses a combination of DC / DC bidirectional modules and AC / DC bidirectional modules to connect to the power grid; Architecture 3 is a DC bus coupling type, which uses all DC / DC bidirectional modules and external backup battery packs to achieve grid-free operation.

[0024] The charge / discharge test range can achieve three working modes: interchangeable charge / discharge, synchronous charging, and synchronous discharging. Interchangeable charge / discharge involves discharging one group of batteries while charging another; synchronous charging and discharging involve multiple groups of batteries simultaneously performing the same type of step. The test procedure documents are compatible with lead-acid, lithium-ion, and nickel-metal hydride batteries. Discharging includes constant current, constant power, constant resistance, and programmable discharge modes, while charging includes constant current multi-stage and constant power multi-stage modes. Constant current charging is further divided into three standard stages: constant current voltage limiting, constant voltage current limiting, and float charging.

[0025] Energy transfer processes must account for bidirectional module conversion losses and DC bus line losses. The capacity configuration rules and State of Charge (SOC) management rules for backup battery packs are as follows: Capacity configuration rules: There are two types: standard configuration and high-capacity configuration. In the standard configuration, the backup battery stores 50%-70% of its own rated capacity. This configuration only supports the two sets of batteries to be interchanged for charging and discharging, and cannot be charged or discharged synchronously. The high-capacity configuration requires that the rated capacity of the backup battery is not less than 3 times the rated capacity of a single set of batteries under test. It can be compatible with the three operating modes of interchangeable charging and discharging, synchronous charging, and synchronous discharging.

[0026] State of Charge (SOC) management rules: Two protection thresholds are set, with an upper limit of 95% and a lower limit of 20%. When the backup battery's SOC reaches 95%, the system stops feeding excess energy back to it to prevent overcharging damage. When the backup battery's SOC drops to 20%, the system stops feeding energy back from it to prevent over-discharge damage. During system initialization, the backup battery is pre-charged to 30% SOC by default, enabling it to both absorb and release energy.

[0027] The driving logic of the bidirectional converter modules under each architecture is permanently stored in the local driver mapping table. The input is the current mutual power demand, and the output is the duty cycle of the corresponding topology's switching transistors or the relay action sequence, ensuring that the energy transmission path is unique and safe. Architecture 1 relies on relay timing logic to avoid short-circuit risks. Architecture 2 stabilizes the DC bus voltage through closed-loop control and achieves grid-connected synchronous frequency, phase, and amplitude. Architecture 3 performs float charging or intermittent supplementary charging control for the backup battery according to the operating conditions.

[0028] When the system starts, it reads the backup battery capacity configuration parameters: if it is a standard configuration (50%-70% of rated capacity), synchronous charging and synchronous discharging modes are prohibited, and only interchangeable charging and discharging is allowed; if it is a high-capacity configuration (≥3 times the rated capacity), synchronous mode is allowed.

[0029] Furthermore, the process of obtaining the task feature parameters is as follows: A comprehensive data acquisition link was established to synchronously collect multi-dimensional raw operating condition data from multiple sets of batteries under test, bidirectional conversion modules, and test stations that have been connected to the charge-discharge test site. Data preprocessing was performed on the multi-dimensional raw operating condition data, and dimension alignment, unit unification, and feature fusion were carried out on the cleaned multi-dimensional raw operating condition data to construct a structured feature dataset and obtain task feature parameters.

[0030] like Figure 2 As shown, the established full-domain data acquisition link covers all measurement points of the battery body, power conversion equipment, bus circuit, and buffer unit. It synchronously collects multi-dimensional raw operating condition data, specifically including core operational information such as battery type, rated capacity, real-time SOC, preset charge / discharge mode, bidirectional module model and type, system topology, energy buffer unit type, real-time conversion loss of bidirectional modules, and DC bus line loss.

[0031] The data preprocessing stage sequentially performs threshold removal, noise suppression, and abnormal data interception operations. It filters invalid operating condition data such as battery overcharging, over-discharging, device failure, and line abnormality through preset threshold discrimination logic, and retains valid original data samples.

[0032] After data cleaning, dimensional alignment and unit unification are performed on various heterogeneous data, converting different physical quantities such as power, energy, state of charge, and loss coefficient into industry-standard dimensionless values. Feature fusion adopts a normalized splicing method, integrating battery parameters, equipment parameters, architecture parameters, and process parameters in sequence according to a preset fixed sorting rule, and uniformly splicing them to generate a 128-dimensional standardized feature vector, which is the final task feature parameter.

[0033] Furthermore, the process of obtaining the optimal scheduling strategy is as follows: The task feature parameters are used as a multi-dimensional retrieval index. The similarity matching algorithm is called to traverse the preset historical working condition knowledge base, compare the current working condition with the historical working condition, calculate the feature matching score in combination with the working condition coupling degree judgment standard, compare the feature matching score with the preset matching threshold, and filter out the set of the same source historical operation schemes that are higher than the matching threshold. Within the set of the same source historical operation schemes, the scheme with the highest feature matching score is selected as the optimal scheduling strategy.

[0034] The historical operating condition knowledge base is stored in partitions according to hardware architecture, battery type, backup battery configuration, and charging / discharging mode. Each historical data entry is bound to a corresponding 128-dimensional feature vector, historical scheduling scheme, energy operation record, and feedback rate data. Using the 128-dimensional task feature vector generated earlier as a multi-dimensional retrieval index, a weighted similarity matching algorithm is called to traverse the historical operating condition knowledge base line by line, completing a full comparison between the current operating condition and each historical operating condition in the database, and outputting the original comparison result for each historical operating condition.

[0035] The specific criterion for determining the degree of coupling under operating conditions uses the phase velocity calculation formula to determine the feature matching score. The specific calculation formula is as follows: ; in, This represents the similarity of working conditions, or feature matching score. The higher the value, the better the match between the current working condition and the historical working conditions. This represents the total dimension of the feature parameters, with a value of 128. The value of the i-th feature parameter for the current task; The value of the i-th feature parameter under historical operating conditions; The i-th dimension feature weight coefficient is used to assign higher weights to core dimensions such as battery capacity, system architecture, energy buffer type, and backup battery configuration.

[0036] By similarity As the basis for determining operating condition coupling, the system pre-configures a globally unified matching threshold. The calculated feature matching score is compared with... Perform numerical comparison and retain only ≥ Historical operating schemes were used to select a set of historical operating schemes that closely match the operating conditions.

[0037] In the set of historical operation schemes from the same source, the feature matching scores of all schemes are traversed, and the historical operation scheme with the highest score is selected and directly determined as the optimal scheduling strategy.

[0038] Furthermore, the process of obtaining the initial running strategy is as follows: The system collects environmental parameters from the entire charging and discharging test site as constraints for operating condition correction. Using the optimal scheduling strategy as the baseline scheme, it adapts and adjusts the baseline scheme by combining the current operating condition characteristics through an adaptive correction model. The feasibility and adaptability of the adjusted scheme are verified by checking the power outage of the operating condition conflict, and an initial operating strategy adapted to the current operating condition characteristics is output.

[0039] The overall environmental parameters include DC bus voltage, bidirectional module operating status, grid voltage and frequency parameters, and real-time SOC of the backup battery. These parameters serve as external constraints for the adaptive correction model. The adaptive correction model uses the optimal scheduling strategy as the baseline scheme and combines system grid connection rules, backup battery state-of-charge protection rules, and equipment rated parameters to complete comprehensive adaptation and adjustment.

[0040] The specific details of the adaptive correction model are as follows: The power configuration, working step sequence, and energy compensation threshold of the baseline scheme are combined with the real-time backup battery SOC, grid connection parameters, and DC bus status and input into the adaptive correction model. The adaptive correction model calculates the power correction coefficient based on the real-time SOC of the backup battery. When the SOC is greater than 90%, the upper limit of the discharge side power is reduced by 5% to reduce the flow of electrical energy to the backup battery. When the SOC is less than 30%, the lower limit of the charging side power is increased by 3% to ensure reasonable discharge and energy replenishment of the backup battery. When the SOC is in the normal range of 30%-90%, the power parameters of the baseline scheme remain unchanged. For both AC bus and DC bus grid connection architectures, the grid frequency, phase and amplitude conditions are checked synchronously. If the grid parameters deviate slightly, the bidirectional module drive parameters are finely adjusted to ensure grid connection stability. Based on the real-time DC bus voltage, the duty cycle of each bidirectional module is corrected to maintain the bus voltage within the rated operating range. After completing the multi-dimensional parameter adaptation, a preliminary adjustment plan is output.

[0041] Operating condition conflict verification process: For the initial adjustment plan, conduct operating condition conflict detection, and check the battery charging and discharging voltage / current range, bidirectional module operating range, and backup battery SOC protection range item by item; if the plan parameters cause the battery to show an overcharge or over-discharge trend, or the backup battery is about to reach the 20% lower limit or 95% upper limit protection threshold, then an operating condition conflict is determined to exist, and the plan is returned to the adaptive correction model for iterative adjustment; if there is no operating condition conflict, then proceed to the power over-limit verification stage.

[0042] Power Exceedance Verification Execution Process: For schemes that pass the operating condition conflict verification, check parameters such as rated charging and discharging power, grid / backup battery compensation power, and module output power; compare all operating power with the rated power on the equipment nameplate; if any power exceeds the limit, it is determined to be a power exceedance, and the model is returned for secondary correction; if all power parameters are lower than the equipment's rated limits, the scheme is deemed legal and valid. After both verifications pass, an initial operating strategy adapted to the current operating condition characteristics is output.

[0043] Furthermore, the specific details of the test timing planning scheme are as follows: The standardized test process files bound to each battery pack under test are retrieved. Through semantic parsing and parameter decomposition, the core timing parameters corresponding to each test step are separated. Using the initial running strategy as the constraint boundary and combined with the battery charging and discharging characteristic deduction algorithm, the power change trend under different step switching nodes is predicted. The power change trend is input into the greedy algorithm, and the timing solution is performed with the goal of maximizing the energy mutual feedback rate in the field. Based on the solution results, a test timing planning scheme is generated.

[0044] The system reads the standardized test procedure file corresponding to each group of batteries under test one by one. This file can be adapted to various conventional test steps such as constant current discharge, constant power discharge, constant resistance discharge, programmable discharge, constant current multi-stage charging, and constant power multi-stage charging.

[0045] According to the established format and control logic of the document, the step identifier, running instructions, and threshold limit content are identified. The parameters are then split, classified, and formatted in sequence. Comments and invalid configuration items in the document are removed, and the core timing parameters corresponding to each test step are extracted.

[0046] The extracted parameters include the start time of a single step, the end time, the step switching trigger condition, the rated operating voltage, the rated operating current, the rated operating power, the stage switching voltage threshold, and the stage switching current threshold, covering the time nodes and electrical constraints of the entire charging and discharging process.

[0047] Using the initial operating strategy as the overall constraint and combining it with core timing parameters, this scheme predicts the trend of power variation in each operation step. The scheme employs a second-order RC equivalent circuit (Thevenin model) to perform characteristic derivation. The corresponding model parameters are retrieved from the middle.

[0048] The specific prediction process is as follows: Based on the rated current and voltage of each step, and relying on the second-order RC equivalent circuit, calculate the variation law of battery terminal voltage and dynamic internal resistance during different operating periods; according to the voltage and current step switching conditions preset in the document, determine the operating state at the moment of step switching: when the operating parameters reach the switching threshold, the battery will experience voltage jumps and current surges. Combining the inherent electrical characteristics of the battery, determine the instantaneous voltage and current values ​​at the moment of switching; after completing the calculation of the battery-side electrical parameters, introduce the inherent loss parameters of the system to carry out power correction. The loss parameters include the fixed conversion loss of the bidirectional converter module and the transmission loss of the DC bus line. Both types of losses are calculated according to the factory calibration values ​​of the equipment. Based on the current and voltage values, calculate the power loss of a single bidirectional module and the power loss of the bus line respectively. Superimpose the battery output / input power with the system power loss to obtain the overall real-time power; according to the time sequence, calculate the real-time power values ​​of each step steady-state operation stage and each step switching node sequentially, and connect the discrete power data in series according to the operating sequence to form a complete and continuous power change trend throughout the entire process.

[0049] The second-order RC equivalent circuit (Thevenin model) is a commonly used equivalent model for the dynamic characteristics of batteries, consisting of an open-circuit voltage source and an ohmic internal resistance. First-stage RC polarization branch ( , ), Second-stage RC polarization branch ( , The battery is connected in series. Let the real-time operating current of the battery be I, and t be the operating time. The formula for calculating the real-time terminal voltage U of the battery is: ; in, The polarization voltage of the first-stage RC branch is calculated using the following formula: ; The polarization voltage of the second-stage RC branch is calculated using the following formula: ; The open-circuit voltage of the battery is determined by the battery type and real-time SOC. , These are the primary polarization resistance and polarization capacitance, respectively, representing the short-time electrochemical polarization characteristics; , These are the second-order polarization resistor and polarization capacitor, respectively, representing the long-term concentration polarization characteristics.

[0050] Steady-state operation phase of the process: Substitute the rated current and duration of the process into the formula, continuously calculate the terminal voltage at different times, and obtain the voltage change pattern; During process switching: taking advantage of the characteristic that capacitor voltage cannot change abruptly, the voltage before switching is retained. , The numerical values ​​are updated only for the operating current I, and the instantaneous terminal voltage during switching is calculated to obtain the electrical parameter jump values. Power conversion: Combining the calculated terminal voltage U and operating current I, the battery-side power P=UI is obtained. Then, the bidirectional converter module loss and DC bus line loss are added to obtain the total real-time power of the system.

[0051] The local preset parameter table is divided into two main categories: battery type and rated capacity. Battery type is differentiated into lead-acid, lithium, and nickel-metal hydride batteries, and batteries of the same type are further divided into different tiers based on rated capacity. Each entry in the table corresponds to a complete set of model parameters. The system first identifies the model, type, and rated capacity of the battery under test, searches the local parameter table according to the above classification rules, and after matching the corresponding entry, directly reads the complete set of parameters, substitutes them into the second-order RC equivalent circuit, and performs power change trend calculation.

[0052] The overall power change trend and initial operating strategy are used as the basis for calculation to carry out timing scheduling. This step adopts a constrained greedy scheduling method, with energy feedback rate as the core evaluation indicator. Two restrictions are set during operation: first, the instantaneous total power of the system must not exceed the rated capacity of the DC bus; second, the number of high-power charging tasks executed simultaneously is controlled. 80% of the rated charging power of a single battery group is set as the high-power judgment standard.

[0053] The specific process for orchestration and iteration based on energy feedback rate is as follows: Based on the inter-group mutual feedback energy per unit time, high-power discharge steps and high-power charging steps are prioritized for synchronous operation, and the execution order and start-up time of each step are initially determined. The total power of all charging tasks within the same time period is calculated. If it exceeds the high power judgment standard, the start time of low priority tasks is postponed. At the same time, the instantaneous total power of the system is checked to ensure that it does not exceed the rated capacity of the DC bus and to avoid illegal operating conditions. According to the current arrangement plan, calculate the corresponding energy mutual feedback rate value, compare the mutual feedback rate results under different time sequence combinations, retain the arrangement with higher mutual feedback rate, and for the period with low mutual feedback rate, fine-tune the start, pause and switch times of the corresponding work steps, shorten the operating range of the discharge work step and the charging work step, increase the proportion of direct energy transmission between groups, and reduce the energy intervention of the grid or backup battery. Constraint verification and mutual feedback rate comparison and adjustment are repeatedly performed until the energy mutual feedback rate under the current layout scheme reaches the optimal level under this working condition, and all operating parameters meet the equipment and process requirements. Finally, the timeline of each group of batteries, task priority, and power linkage relationship are integrated to form a global test timing planning scheme that can be implemented.

[0054] Furthermore, the dynamic power balancing process is as follows: Based on the initial operation strategy and test timing plan, the bidirectional conversion modules of each group are driven in a coordinated manner to carry out closed-loop energy mutual feedback transmission between multiple battery groups; real-time power data of the energy transmission link is collected, the power deviation on both sides of charging and discharging is calculated, and the supply and demand status of the system is determined; the hierarchical equalization control logic is invoked, and the corresponding adjustment level is matched according to the power deviation to control the energy buffer unit to perform corresponding operations, so as to achieve dynamic power balance.

[0055] like Figure 3 As shown, the system uniformly issues control commands according to the initial operation strategy and test timing plan, driving all bidirectional conversion modules to operate in an orderly manner. Combining three types of hardware architectures—AC bus coupling, DC bus coupling, and backup battery buffer—and working modes such as interchangeable charging and discharging, synchronous charging, and synchronous discharging, it builds a closed-loop energy transmission circuit between multiple battery groups to complete inter-group energy mutual feedback.

[0056] Throughout the energy feedback operation, the real-time operating power of the discharge circuit, charging circuit, DC bus, and power conversion unit is continuously collected, and the total output power on the discharge side is obtained by summing the data. Total input power on the charging side .

[0057] The formula for calculating power deviation is as follows: ; in, This represents the system power deviation.

[0058] Determine the system's supply and demand status based on the calculation results: when When the system has a power surplus, the output energy on the discharge side is greater than the energy required on the charging side; when When the system is deemed to have a power deficit, the output energy on the discharge side cannot meet the usage requirements of the charging side; when When the value is 0, the system's energy supply and demand are considered balanced, and no additional adjustment is required.

[0059] The system has three preset control levels and corresponding actions, based on the absolute value of the power deviation. Match the corresponding gear and activate the graded balance control logic: First gear: ≤200W is considered a minor power deviation. Only the duty cycle of the corresponding DC / DC bidirectional converter module's switching transistor is slightly adjusted to change the module's output / input power, gradually offsetting the power deviation without altering the existing test tasks and timing.

[0060] Second gear: 200W ≤500W is considered a medium power deviation. An external energy buffer unit is activated for regulation: If the system is connected to the grid, excess power is transferred to the grid when there is a power surplus, and the grid supplements the difference when there is a power deficit; if the system uses a backup battery pack as a buffer unit, the backup battery pack is charged and stored when there is a power surplus, and the backup battery pack releases power to supplement the deficit when there is a power deficit. The regulation process strictly follows the backup battery SOC management rules: charging stops when the SOC reaches 95%, and discharging stops when the SOC drops to 20%.

[0061] Level 3 gear: A power deviation greater than 500W is considered a significant power deviation. Based on the second-level adjustment, and according to task priority, the lower-priority charge / discharge test tasks are forcibly paused to reduce the overall system power load and quickly equalize the power difference between the charge and discharge sides.

[0062] The entire set of graded control actions is continuously executed until the power deviation returns to the balance range, ultimately achieving dynamic power balance across the entire link and ensuring stable system operation.

[0063] Furthermore, the process of obtaining the test resource scheduling instruction is as follows: Under the operating condition of dynamic power balance, energy operation data of the charging and discharging test site are continuously collected. The mutual feedback rate is calculated according to the mutual feedback rate calculation formula. The mutual feedback rate is connected to the hierarchical scheduling rule base. The operating condition level is divided according to the deviation of the index and the duration of the abnormality. Test resource scheduling instructions are generated according to the operating condition level.

[0064] This scheme uses a time-segmented cumulative method to calculate the full-cycle energy feedback rate. This indicator is the cumulative value of a single complete test cycle and is not used for instantaneous operating condition calculations. Within the same operating period, the system only exists in one state: power deficit or power surplus. Grid-supplied energy and grid-feedback energy will not be generated simultaneously in a single time period; they are accumulated over the entire cycle before being included in the calculation. The formula for calculating the feedback rate is as follows: ; in, Contribution rate of internal energy feedback throughout the entire cycle. This refers to the total cumulative energy transferred between the battery packs within the test site. The cumulative value of the difference in energy supplied to the system by the power grid or backup batteries throughout the entire cycle (generated during power gap periods); It is the cumulative value of excess energy transferred out of the system throughout the entire cycle, including two parts: surplus energy stored in the on-site backup battery and surplus energy fed back to the grid (generated during periods of power surplus).

[0065] When the system is in dynamic power balance operation, it continuously collects energy operation data for the entire test site, specifically including inter-group mutual feedback energy within the test site, differential energy supplemented by the grid or backup batteries, excess energy fed back to the grid or backup batteries, and total test energy flow across the entire site. Based on the mutual feedback rate calculation formula, the collected real-time data is calculated to obtain the energy mutual feedback rate value at the current moment.

[0066] The system compares the real-time feedback rate with the system's preset standard reference feedback rate to calculate the rate of decline in the indicator, and records the duration of the abnormal state. Both parameters are simultaneously transmitted to the hierarchical scheduling rule base to complete the condition level determination. The condition level classification and corresponding handling procedures are as follows: Level 1 Operating Condition (Slight Imbalance) Judgment Criteria: The mutual feedback rate decreases by 5%-10% relative to the standard value, and the duration of this abnormal state does not exceed 30 seconds. Handling Process: Assuming a minor energy imbalance, no changes to the overall task architecture and timing plan are required. Only fine-tuning instructions are issued for the charging / discharging steps where deviations occur, slightly correcting the operating power of individual battery banks and the timing of step switching to gradually restore the energy mutual feedback level. After the instructions are issued, continuous monitoring of operating condition changes continues.

[0067] Level 2 operating condition (moderate imbalance) judgment criteria: The feedback rate decreases by 10%~20% relative to the standard value, and the duration of this abnormal state is greater than 30 seconds and no more than 2 minutes. Handling process: The system is judged to have a moderate energy imbalance, and local adjustments in the work steps cannot improve the situation. The system generates task reorganization scheduling instructions to adjust the access timing and overall startup sequence of each group of batteries under test, rearrange the task layout within the field, optimize the energy matching relationship between groups, and reduce the energy intervention ratio of external buffer units.

[0068] Level 3 Operating Condition (Severe Imbalance) Judgment Criteria: The feedback rate decreases by more than 20% relative to the standard value, and the duration of this abnormal state is greater than 2 minutes. Handling Process: The system is judged to have a severe energy imbalance, and the existing operating strategy and timing scheme are no longer suitable for the current operating condition. The system generates a global reconfiguration scheduling instruction, triggering a process backtracking mechanism. It then re-invokes the historical operating condition knowledge base to match the optimal scheduling strategy and, combined with the on-site operating conditions, recalculates and generates a test timing plan, correcting the operating mode from a top-level logical perspective.

[0069] All scheduling instructions are standardized and encapsulated according to the system's unified communication protocol and sent to each control unit in the charge and discharge test field for execution. During the execution process, energy data and feedback rate are monitored synchronously throughout, forming a closed-loop control system of data acquisition, level determination, instruction issuance, and status feedback.

[0070] Furthermore, the method also includes system iterative optimization, the specific process of which is shown below: After all the battery packs under test have completed the full charge and discharge cycle, the operating condition information, feedback rate, and iterative operation strategy adjusted by instructions are collected to form the complete operation dataset for this cycle. After the complete operation dataset for this cycle is structured, labeled with features, and validated, it is written into the historical operating condition knowledge base according to the knowledge base storage specifications to complete the iterative update of the historical operating condition knowledge base. After the knowledge base is updated, the system automatically jumps to the start of the new test task.

[0071] like Figure 4 As shown, after all the battery packs under test have completed the entire charging and discharging operation, the system will collect all the operation data throughout the process. Specifically, this includes power loss data of the bidirectional converter module, DC bus line loss data, energy buffer unit operation status data, voltage, current, and state of charge data of each battery pack, etc. At the same time, the system will calculate the average mutual feedback rate of the entire cycle, record the final iterative operation strategy after dynamic adjustment by on-site dispatch instructions, and integrate all the information into a complete operation dataset for this cycle.

[0072] The system performs structured compilation of the dataset according to preset classification rules, completes feature annotation by combining the battery parameters, system architecture, and operating mode corresponding to the current task, and associates the dataset with the corresponding operating condition labels. Before data is stored, validity verification is performed, comparing the average feedback rate of the current round with the historical average feedback rate of the same type of operating condition. If the feedback rate of the current round is 30% or more lower than the historical average, or if there are abnormal operation records such as abnormal power outages, emergency shutdowns, or manual intervention in the data, the dataset is directly marked as invalid data and is prohibited from being written into the historical operating condition knowledge base. This prevents abnormal data from polluting the inventory samples and avoids the problem of reference bias in operating strategies.

[0073] Only valid datasets that have passed verification are entered into the database in strict accordance with the knowledge base partitioning and storage specifications. They are then classified and archived according to system architecture, battery type, backup battery capacity configuration, and charging / discharging mode to complete the iterative update of the historical operating condition knowledge base.

[0074] After the knowledge base is updated, the system automatically switches to the new round of test task initiation state. The strategy reuse mapping rules are as follows: the optimal scheduling strategy verified in this round, including the complete set of content such as the execution sequence of work steps, the charging and discharging power ratio, and the start-up and calling rules of the energy buffer unit, is directly used as the initial benchmark strategy for the next round of the same type of working condition; if the characteristic parameters and field environment of the next round of test task change, the system automatically calls the adaptive correction model mentioned above, and makes parameter fine-tuning of the benchmark strategy in combination with the real-time working conditions to generate an initial running strategy adapted to the new scenario.

[0075] The processes described above with reference to the flowcharts in the embodiments disclosed in this invention can be implemented as computer software programs. The embodiments disclosed in this invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), it performs the functions defined in the methods of this application. It should be noted that the computer-readable medium described above in this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wire segments, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless segments, wire segments, optical fibers, RF, etc., or any suitable combination thereof.

[0076] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. 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 a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated 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.

[0077] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are merely examples and do not limit the present invention. The purpose of the present invention has been fully and effectively achieved. The functions and structural principles of the present invention have been shown and explained in the embodiments. Without departing from the stated principles, the implementation of the present invention may have any variations or modifications.

Claims

1. A bidirectional charging and discharging method for a battery pack based on energy mutual feedback, characterized in that, The method includes: Multiple sets of battery packs under test are connected to a charge-discharge test field equipped with a bidirectional conversion module and an energy buffer unit to extract task characteristic parameters. Based on task feature parameters and a pre-set historical operating condition knowledge base, the optimal scheduling strategy is matched and obtained. The optimal scheduling strategy is then corrected based on the initial operating state of the test field to obtain the initial operating strategy. Retrieve the test process files bound to each battery pack under test, parse the power and time parameters of each step in the test process files, combine the initial running strategy, predict the power change trend corresponding to each step, and generate a test timing plan through a greedy algorithm. Based on the initial operation strategy and test timing plan, the bidirectional conversion module is driven to work, and the power difference of energy transmission is monitored in real time. According to the power difference status, the energy buffer unit performs dynamic power balancing. During charging and discharging operation, energy operation data is collected in real time. The energy operation data is calculated using a preset mutual feedback rate calculation formula to obtain the mutual feedback rate. Combined with a preset scheduling rule base, test resource scheduling instructions are generated to adjust the test task operation status.

2. The bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 1, characterized in that, The process of obtaining the task feature parameters is as follows: Establish a full-domain data acquisition link to synchronously collect multi-dimensional raw operating condition data from multiple sets of batteries under test, bidirectional conversion modules, and test stations that have been connected to the charge and discharge test site. Data preprocessing is performed on the multi-dimensional raw working condition data. The cleaned multi-dimensional raw working condition data is then aligned in dimensions, standardized in units, and fused in features to construct a structured feature dataset and obtain task feature parameters.

3. The bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 2, characterized in that, The process of obtaining the optimal scheduling strategy is as follows: The task feature parameters are used as a multi-dimensional retrieval index. The similarity matching algorithm is called to traverse the preset historical working condition knowledge base, compare the current working condition with the historical working condition, calculate the feature matching score in combination with the working condition coupling degree judgment standard, compare the feature matching score with the preset matching threshold, and filter out the set of the same source historical operation schemes that are higher than the matching threshold. Within the set of historical operation schemes from the same source, the scheme with the highest feature matching score is selected as the optimal scheduling strategy.

4. The bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 3, characterized in that, The process of obtaining the initial running strategy is as follows: The environmental parameters of the entire charging and discharging test site are collected as constraints for the working condition correction. The optimal scheduling strategy is used as the baseline scheme. The baseline scheme is adapted and adjusted by an adaptive correction model in combination with the current working condition characteristics. The feasibility and adaptability of the adjusted scheme are verified by checking for power outages due to operating conditions, and an initial operating strategy adapted to the characteristics of the current operating conditions is output.

5. A bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 4, characterized in that, The specific details of the test timing planning scheme are as follows: The standardized test process files bound to each battery pack under test are retrieved, and the core timing parameters corresponding to each test step are separated through semantic parsing and parameter decomposition. Using the initial operating strategy as a constraint boundary and combining the battery charging and discharging characteristics deduction algorithm, the power change trend under different work step switching nodes is predicted; The power change trend is input into a greedy algorithm, and the timing solution is performed with the goal of maximizing the energy mutual feedback rate in the field. Based on the solution results, a test timing planning scheme is generated.

6. The bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 5, characterized in that, The dynamic power balancing process is as follows: Based on the initial operation strategy and test timing plan, the bidirectional conversion modules of each group are driven in a coordinated manner to carry out closed-loop energy mutual feed transmission between multiple battery groups. Collect real-time power data of the energy transmission link, calculate the power deviation on both sides of the charge and discharge, and determine the supply and demand status of the system; The hierarchical balancing control logic is invoked, and the corresponding adjustment level is matched according to the power deviation to control the energy buffer unit to perform corresponding operations, thereby achieving dynamic power balancing.

7. A bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 6, characterized in that, The process of obtaining the test resource scheduling instruction is as follows: Under the operating condition of dynamic power balance, the energy operation data of the charge and discharge test field are continuously collected, and the mutual feedback rate is calculated according to the mutual feedback rate calculation formula. The feedback rate is integrated into the hierarchical scheduling rule base. The operating condition level is divided based on the deviation of the indicator and the duration of the abnormality. Test resource scheduling instructions are generated according to the operating condition level.

8. A bidirectional charging and discharging method for a battery pack based on energy mutual feedback according to claim 7, characterized in that, The method also includes system iterative optimization, the specific process of which is shown below: After all the battery packs under test have completed the full charge and discharge cycle, the operating condition information, the feedback rate throughout the cycle, and the iterative operation strategy adjusted by instructions are collected to form the complete operation dataset for this cycle. After the complete dataset of this round of operation is structured, labeled with features and validated, it is written into the historical working condition knowledge base according to the knowledge base storage specifications to complete the iterative update of the historical working condition knowledge base; Once the knowledge base is updated, you will be automatically redirected to the next test task launch phase.

9. A bidirectional charging and discharging system for a battery pack based on energy mutual feedback, characterized in that, The system is used to execute a bidirectional charging and discharging method for a battery pack based on energy mutual feedback, as described in any one of claims 1-8.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is executed by a processor to implement a bidirectional charging and discharging method for a battery pack based on energy mutual feedback as described in any one of claims 1-8.

Images