A battery equalization control method and system based on dynamic pressure difference

Abstract

This invention belongs to the field of battery equalization control technology, specifically relating to a battery equalization control method and system based on dynamic differential pressure. The method includes: collecting the voltage, temperature, and current of individual batteries; constructing a multi-dimensional state potential index that includes factors such as state of charge, temperature suitability, and health status; and dynamically dividing the energy supply set and receiving set accordingly. The system also possesses a transient overvoltage identification and mitigation mechanism, capable of quickly locating overvoltage points and finding the adjacent safe battery with the lowest state of charge as the mitigation target. During equalization, the algorithm comprehensively considers the ohmic internal resistance and health status of both the supply and demand sides to calculate the current limit value, and uses a PID algorithm to dynamically adjust the duty cycle for energy transfer. When the standard deviation, rate of change, and equalization current of the state potential index tend to stabilize, the system automatically terminates the equalization process. This invention improves equalization efficiency and prevents the risk of overcharging or thermal runaway.

Description

Technical Field

[0001] This invention relates to the field of battery balancing control technology. More specifically, this invention relates to a battery balancing control method and system based on dynamic differential pressure. Background Technology

[0002] Battery banks are used in communication base stations, data centers, and energy storage systems, primarily to supply power to critical equipment after a mains power outage. To extend the lifespan of battery banks, it is typically necessary to monitor the operating parameters of individual cells in real time, such as temperature, voltage, and internal resistance, and to perform routine management and maintenance on individual cells that exceed the float charge voltage range based on the monitoring data.

[0003] In existing technologies, the common approach for maintaining the voltage balancing of individual cells in a battery pack is as follows: when the voltage of an individual cell is higher than the float charge voltage by a certain range, the battery pack performs discharge balancing through a resistive load, dissipating electrical energy as heat; when the voltage of an individual cell is lower than the float charge voltage by a certain range, an external power source charges the battery pack for balancing. However, this type of approach generally suffers from problems such as delayed maintenance response, the need for external power intervention, and significant energy loss. Furthermore, current management and maintenance largely rely on manual operation or passive reminders based on preset time points, lacking the ability to deeply analyze and finely adjust the internal electrochemical potential of the battery under complex dynamic operating conditions.

[0004] Referring to Chinese patent document CN108702005B, an online balancing method for battery packs is disclosed. This method collects the voltage of individual battery cells, selects cells that need to be balanced by charging or discharging, obtains the deviation between the individual cell voltage and the float charge voltage, and determines the balancing parameters based on the deviation to perform balancing processing on the cells. This achieves automated online balancing maintenance without manual intervention.

[0005] While the aforementioned existing technologies have achieved a certain degree of automated maintenance, some problems still exist when facing dynamic operating environments such as high-rate charging and discharging. Equalization judgment based on the deviation between individual cell voltage and float charge voltage cannot comprehensively assess the battery's state of charge, temperature deviation, and true health status under operating conditions, leading to limitations in equalization judgment and a tendency to cause erroneous equalization. Secondly, during energy transfer, the electrical topology distance, internal ohmic resistance, and health tolerance of the supply and demand sides are not assessed, and there is a lack of thermodynamic constraints on the equalization current. This can easily cause secondary damage to fragile cells due to excessive transfer current, and it cannot effectively solve the problem of local transient overvoltage. Finally, the lack of multi-dimensional joint feedback judgment logic for equalization convergence can easily lead to repeated oscillations or prolonged idling during the equalization process. Summary of the Invention

[0006] To address the technical problems in existing technologies, such as inaccurate and misjudged battery status assessments due to reliance on voltage or a single capacity index for equalization determination, and the lack of aging constraints on internal resistance and health status and rapid response mechanisms for local overvoltage during energy transfer, which lead to the risk of thermal runaway or inefficient equalization processes and the potential for secondary damage to individual cells, this invention provides solutions in the following aspects.

[0007] In a first aspect, the present invention provides a battery equalization control method based on dynamic differential pressure, comprising: S1, collecting the voltage, temperature and charge / discharge current of individual cells in a battery pack, measuring the ohmic internal resistance of individual cells, obtaining the state of charge, temperature deviation and health state factor based on the temperature and the charge / discharge current and constructing a state potential index, classifying individual cells with the state potential index greater than a first threshold into an energy supply set, and classifying individual cells with the state potential index less than a second threshold among those that meet preset safety conditions into an energy receiving set; S2. Identify individual cells whose voltage exceeds a preset overvoltage threshold as transient overvoltage points. Among the individual cells adjacent to each other in the switch control circuit that meet the preset safety conditions, select the one with the lowest state of charge as the target for facilitating transient overvoltage points. S3. Take the energy supply set and energy receiving set or transient overvoltage point and diversion target as the supply and demand sides, calculate the current limit value based on the ohmic internal resistance and health status factor of the supply and demand sides, and input the real-time voltage difference and current limit value into the proportional integral derivative control algorithm to determine the duty cycle and perform balancing. S4. The process terminates when the standard deviation and rate of change of the state potential index are lower than the first convergence threshold and the second convergence threshold, respectively, and the equilibrium current is stable.

[0008] This invention objectively assesses the true state of a battery by constructing a state potential index from multiple dimensions, avoiding misjudgments caused by a single indicator. At the same time, it establishes a dual mechanism of global supply and demand balance and local transient overvoltage mitigation, which not only improves the consistency of the overall energy distribution but also quickly eliminates the safety hazards of local overvoltage. It introduces internal resistance and health state factors to constrain the balance current and prevent excessive current from causing secondary damage to the battery.

[0009] Preferably, the step of obtaining the state of charge, temperature deviation, and health status factor based on the temperature and the charge / discharge current, and constructing a state potential index, includes: obtaining the state of charge based on the charge / discharge current using the ampere-hour integration method; using the deviation value between the temperature and a preset standard operating temperature range as the temperature deviation and converting it into a temperature suitability deduction factor; calculating the attenuation rate using the cumulative number of charge / discharge cycles and the number of reference cycles, taking the difference between 1 and the attenuation rate and comparing it with a preset minimum positive value, and taking the larger one as the health status factor; and obtaining the state potential index by weighted summation of the state of charge, the temperature suitability deduction factor, and the health status factor, wherein the weight coefficient of the temperature suitability deduction factor is negative.

[0010] This invention quantifies and integrates charge / discharge capacity, temperature deviation, and cycle life degradation, accurately reflecting the true energy transmission and reception potential and health tolerance of a single battery cell under current operating conditions, thus improving the scientific nature of energy supply and demand allocation.

[0011] Preferably, the preset safety conditions specifically include: the temperature deviation is less than a preset temperature limit value and the health state factor is greater than a preset safety threshold; obtaining the first threshold and the second threshold includes: calculating the average baseline value of the state potential index of all individual cells; adding the product of the absolute value of the average baseline value and a first floating coefficient to the average baseline value to obtain the first threshold; subtracting the product of the absolute value of the average baseline value and a second floating coefficient from the average baseline value to obtain the second threshold.

[0012] This invention adaptively generates dynamic thresholds based on the average level of the global state potential index. Compared with fixed thresholds, it can better adapt to the differences in battery packs under different aging stages and operating conditions. At the same time, combined with safety condition filtering, it eliminates the risk of deteriorating individual cells being mistakenly selected as energy recipients.

[0013] Preferably, the calculation of the current limit value based on the ohmic internal resistance and health state factor of both the supply and demand sides includes: extracting the smaller value from the health state factors of both the supply and demand sides and multiplying it by a reference safety constant to obtain the initial limit current; according to the relationship: Calculate impedance thermal current limiting ,in For the maximum allowable temperature rise, and These are the ohmic internal resistances of the supply and demand sides, respectively. The equivalent thermal resistance is preset; the minimum value between the initial limiting current and the impedance thermal current limit is taken as the current limit value.

[0014] This invention determines the maximum permissible balancing current through dual constraints, cutting off the inducing paths of high temperature and accelerated aging, and ensuring the safety of balancing execution.

[0015] Preferably, the measurement of the ohmic internal resistance of a single cell includes: injecting a DC test pulse current with a constant amplitude and a pulse width in the millisecond range into the single cell; acquiring the terminal voltage of the single cell as the unperturbed initial voltage before injection; acquiring the terminal voltage of the single cell as the test final voltage at the transient step moment before polarization occurs during injection; and dividing the absolute value of the difference between the test final voltage and the unperturbed initial voltage by the amplitude value of the DC test pulse current to obtain the ohmic internal resistance.

[0016] This invention utilizes millisecond-level high-frequency pulse injection to avoid interference from electrochemical polarization effects on internal resistance measurement, accurately extracting pure ohmic internal resistance parameters, and providing highly reliable hardware parameter support for subsequent thermal current limiting calculations.

[0017] Preferably, identifying a single battery cell whose voltage exceeds a preset overvoltage threshold as a transient overvoltage point includes: when the voltage of the single battery cell exceeds the overvoltage threshold and the duration exceeds a preset anti-shake delay in seconds, determining that the single battery cell has exceeded the limit and marking it as the transient overvoltage point; wherein setting the overvoltage threshold includes: determining whether the operating condition is a charging / discharging state or a static state based on the total current of the main circuit, retrieving the corresponding operating condition compensation parameter and performing an algebraic sum operation with the absolute overvoltage limit value to obtain the overvoltage threshold.

[0018] This invention dynamically adjusts the overvoltage judgment benchmark based on the current charging and discharging state and introduces an anti-shake delay mechanism to filter out the false high voltage caused by high frequency current sudden changes or sampling noise, thereby achieving accurate capture of the real transient overvoltage point.

[0019] Preferably, the step of selecting the cell with the lowest state of charge as the mitigation target from among the adjacent cells in the switch control circuit that meet the preset safety conditions for the transient overvoltage point includes: finding cells adjacent to the transient overvoltage point in the switch control circuit to form a surrounding candidate set; eliminating cells in the surrounding candidate set that do not meet the preset safety conditions; and selecting the cell with the lowest state of charge from the remaining cells as the mitigation target.

[0020] This invention finds the nearest diversion target based on electrical topology, shortens the energy transfer path to reduce transmission heat loss, and ensures that the receiving end has sufficient capacity and no safety hazards, thus achieving efficient release of local overvoltage energy.

[0021] Preferably, the step of inputting the real-time voltage difference and current limit value into the proportional-integral-derivative (PID) control algorithm to determine the duty cycle for balancing includes: dividing the real-time voltage difference by the sum of the ohmic internal resistances of the supply and demand sides to obtain the expected current; taking the smaller value between the expected current and the current limit value as the target current; acquiring the actual balancing current; inputting the deviation between the target current and the actual balancing current into the PID control algorithm to calculate the duty cycle and generate a control command to execute the balancing.

[0022] This invention provides real-time deviation feedback between the calculated target safe current and the actual measured current. By dynamically adjusting the pulse duty cycle, it ensures that the actual energy transfer current always smoothly matches the safety target, thus avoiding current overload.

[0023] Preferably, the process terminates when the standard deviation and rate of change of the state potential index are lower than the first convergence threshold and the second convergence threshold, respectively, and the equilibrium current is stable. This includes: calculating the standard deviation of the state potential index for all individual cells; subtracting the average value of the state potential index of the previous period from the average value of the current period and dividing by the sampling time interval to obtain the rate of change; calculating the discrete variance of the actual equilibrium current data sequence; determining that the equilibrium current is stable when the discrete variance is continuously less than a set variance threshold for a preset sampling period; and cutting off the equilibrium drive signal when the standard deviation is lower than the first convergence threshold, the absolute value of the rate of change is lower than the second convergence threshold, and the equilibrium current is determined to be stable.

[0024] This invention utilizes a multi-condition joint judgment mechanism instead of a single threshold termination, preventing the system from falling into invalid repeated oscillations and reducing system power consumption.

[0025] Secondly, the present invention provides a battery balancing control system based on dynamic differential pressure, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned battery balancing control method based on dynamic differential pressure is implemented.

[0026] By adopting the above technical solution, a computer program for a battery balancing control method based on dynamic pressure difference is generated and stored in a memory so that it can be loaded and executed by a processor. This allows for the creation of a terminal device based on the memory and processor, making it convenient to use.

[0027] The beneficial effects of this invention are as follows: This invention constructs a multidimensional state potential model encompassing ohmic internal resistance, state of charge, temperature deviation, and health factors. Based on this model, a dual control architecture combining global state balancing and local overvoltage mitigation is designed. By introducing thermal current limiting and anti-jitter delay judgment, the safe target current is dynamically calculated and feedback adjustment is performed. This not only improves the balancing efficiency of charge distribution and slows down the aging process of the battery pack, but also blocks the risk of thermal runaway caused by transient overvoltage at local nodes from the source, thereby enhancing the long-term stability and reliability of the battery system under complex operating conditions. Attached Figure Description

[0028] Figure 1 This is a flowchart of a battery balancing control method based on dynamic differential pressure. Figure 2 A scatter plot of the state potential index distribution of a single battery cell; Figure 3 This is a comparison chart showing the maximum monomer pressure difference as a function of the number of cycles. Detailed Implementation

[0029] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0030] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0031] This invention discloses a battery equalization control method based on dynamic voltage difference, referring to... Figure 1 This includes steps S1-S4: S1. Construct the state potential index and divide the dataset.

[0032] In an optional embodiment, the terminal voltage data of individual cells within the battery pack, the temperature signal of the thermistor, and the charging / discharging current data are continuously acquired. When measuring the ohmic internal resistance using the pulse current injection method, the controller controls the bidirectional DC-DC converter to inject a high-frequency square wave DC test pulse current with an amplitude adjusted according to the ambient temperature into the individual cells. The fundamental components of the response voltage and the injected current are extracted using a discrete Fourier transform algorithm. The ratio of the complex phasors of the fundamental components of the response voltage and the injected current is calculated to obtain the complex impedance. The real part of the complex impedance is then extracted to obtain the ohmic internal resistance.

[0033] Based on the unscented Kalman filter algorithm combined with the ampere-hour integral method, the state of charge of a single battery cell is estimated in real time. The absolute value of the difference between the current collected temperature and the preset optimal operating temperature is calculated as the temperature deviation. In this embodiment, the optimal operating temperature is 25℃. Based on the autoregressive moving average model, historical charging and discharging data are statistically analyzed, and the health status factor is obtained by calculating the ratio of the current maximum available capacity to the factory rated capacity.

[0034] The range transformation method is used to normalize the state of charge, temperature deviation, and health state factors. The analytic hierarchy process (AHP) is then used to determine the weighting coefficients for these factors, followed by a weighted summation operation to construct a state potential index. The state potential indices of all individual cells are then sorted in descending order. Using a sequential traversal operation, the device address indices of individual cells whose state potential index values ​​exceed a first threshold are written to the energy supply set linked list, while the device address indices of individual cells that meet preset safety conditions are written to the energy receiving set linked list.

[0035] In an optional embodiment, measuring the ohmic internal resistance of a single cell includes the following steps: A DC test pulse current with a constant amplitude and a preset millisecond pulse width is injected into a single cell. The terminal voltage of the single cell is collected before injection as the unperturbed initial voltage. The terminal voltage of the single cell is collected at the transient step moment before polarization occurs during injection as the test final voltage. The ohmic internal resistance of the single cell is obtained by dividing the absolute value of the difference between the test final voltage and the unperturbed initial voltage by the amplitude value of the DC test pulse current.

[0036] For example, the control pulse generation unit injects a square wave DC test pulse current with a constant amplitude in the range of 5A to 10A into the target single cell, with the pulse width set in the millisecond range of 1ms to 5ms. Within a time window of 5ms to 10ms before the pulse injection, 100 battery terminal voltage data points are continuously collected at a sampling frequency of 10kHz. The arithmetic mean is calculated and stored as the unperturbed initial voltage, set to 3.850V. Within a time interval of 0.1ms to 0.5ms after the DC test pulse current injection, the terminal voltage is collected and recorded as the final test voltage, set to 3.875V, generated by the 10A DC test pulse current injection.

[0037] To accurately obtain the internal ohmic characteristics of a single cell, a relationship was established between the internal ohmic resistance and the final test voltage and the unperturbed initial voltage. The specific relationship is as follows:

[0038] in, Indicates the internal resistance of the ohm. The final voltage is represented by the voltage at which the terminal voltage of a single cell is acquired at the moment of transient voltage step before polarization begins after the injection of a DC test pulse current. This represents the unperturbed initial voltage, obtained by continuously collecting the terminal voltage of individual cells and calculating the arithmetic mean within a preset time period before the pulse injection begins. This indicates the amplitude of the DC test pulse current, which is obtained by setting the current amplitude injected into a single battery cell using the controller.

[0039] The absolute value of the difference between the final voltage and the unperturbed initial voltage is calculated to be 25mV. The absolute value of the difference is then divided by the amplitude of the DC test pulse current, which is 10A, to calculate the ohmic internal resistance of the single cell to be 2.5mΩ.

[0040] In an optional embodiment, the state of charge (SOC), temperature deviation, and health status factor are obtained based on temperature and charge / discharge current, and a state potential index is constructed. This includes the following steps: obtaining the SOC using the ampere-hour integral method based on the charge / discharge current; using the deviation of temperature from a preset standard operating temperature range as the temperature deviation and converting it into a temperature suitability deduction factor; calculating the attenuation rate using the cumulative number of charge / discharge cycles and the number of reference cycles, and comparing the difference between 1 and the attenuation rate with a preset minimum positive value, taking the larger value as the health status factor; and obtaining the state potential index by weighted summation of the SOC, temperature suitability deduction factor, and health status factor, wherein the weighting coefficient of the temperature suitability deduction factor is negative.

[0041] For example, based on the real-time collected charging and discharging current, the ampere-hour integration is performed. Assuming the initial full-charge nominal capacity is set to 50Ah, if the current cumulative discharged capacity reaches 10Ah, the current remaining usable capacity percentage is calculated to be 80%, that is, the state of charge value is equal to 0.8. At the same time, the surface temperature of the individual battery is collected. If the preset standard operating temperature range is set to 25℃ to 35℃, and the current measured temperature is 42℃, the measured temperature exceeds the upper limit boundary by 7℃. Using a conversion factor of 0.02 per degree Celsius, the temperature suitability deduction factor is calculated to be equal to 0.14.

[0042] The cumulative charge-discharge cycle count of a single battery cell is read and set to 400. This cumulative count is divided by the calibrated baseline lifespan of 2000 cycles to calculate the degradation rate, which is 0.2. Subtracting the degradation rate from 1 yields 0.8, which is then compared to a preset minimum positive value of 0.1. The larger value, 0.8, is taken as the health state factor. Weighting coefficients are assigned to each parameter: the weight for the state of charge (SBC) value is set to 0.6, the weight for the health state factor is set to 0.4, and the weight for the temperature suitability deduction factor is set to -0.5. After weighted summation, the state potential index of the single battery cell is obtained as 0.73.

[0043] In an optional embodiment, the preset safety conditions are that the temperature deviation is less than a preset temperature limit and the health state factor is greater than a preset safety threshold. The process involves classifying individual cells with a state potential index greater than a first threshold into an energy supply set, and classifying individual cells meeting the preset safety conditions with a state potential index less than a second threshold into an energy receiving set. This includes the following steps: calculating the average baseline value of the state potential index of all individual cells; adding the absolute value of the average baseline value of the state potential index to the product of the average baseline value of the state potential index and a first floating coefficient to obtain a first threshold, where the first floating coefficient ranges from 0.05 to 0.2; subtracting the absolute value of the average baseline value of the state potential index from the product of the average baseline value of the state potential index and a second floating coefficient to obtain a second threshold, where the second floating coefficient ranges from 0.05 to 0.2; classifying individual cells with a state potential index greater than the first threshold into the energy supply set, and classifying individual cells meeting the preset safety conditions with a state potential index less than the second threshold into the energy receiving set.

[0044] For example, the state potential indices of all 100 individual cells in the battery pack are summed within the same sampling period, and the arithmetic mean is calculated by dividing by the total number of cells. This generates an average baseline value of 0.75 for the state potential index. The product of the absolute value of this average baseline value (0.75) and a first fluctuation coefficient (0.1) is then calculated as a positive tolerance of 0.075. Adding this positive tolerance to the average baseline value yields a first threshold of 0.825.

[0045] The product of the absolute value of the average baseline value of the State Potential Index (SPI), 0.75, and the second floating coefficient, 0.1, is used as the negative tolerance of 0.075. Subtracting the negative tolerance from the average baseline value of the SPI yields a second threshold of 0.675. Cells with an SPI greater than 0.825 are included in the energy supply set; 15 cells meet the first threshold condition and output energy. Cells with a temperature deviation less than a preset temperature limit, a health state factor greater than a preset safety threshold, and an SPI less than 0.675 are included in the energy receiving set; 12 cells meet the condition to receive energy. For example... Figure 2 As shown, this diagram illustrates the distribution of the state potential index of each individual cell within the battery pack and the division of the energy supply and demand set.

[0046] In this way, the ohmic internal resistance of a single cell is accurately extracted by the pulse current injection method. Combined with the state of charge, temperature deviation and health status factors, a multi-dimensional state potential index is constructed, which effectively reduces the performance misjudgment rate caused by relying solely on voltage or capacity assessment. Based on the adaptively generated dynamic threshold, the energy supply set and energy receiving set are accurately divided. Under the premise of ensuring that the battery meets the preset safety conditions, the scientific nature of the balancing object is ensured, and the accuracy of the pre-judgment of energy transfer and the overall stability of the battery system are improved.

[0047] S2. Identify transient overpressure points and select the evacuation target.

[0048] In an optional embodiment, the voltage of individual battery cells is periodically read using a timer interrupt service function, and a sliding window averaging filter algorithm is used to filter out high-frequency sampling noise. The total current of the main circuit is obtained, and the current load condition is determined to be either charging / discharging or idle. Based on the determined condition type, the corresponding condition compensation parameters are retrieved, and the condition compensation parameters are algebraically summed with the absolute overvoltage limit value to calculate and determine the overvoltage threshold. The voltage surge data of each individual battery cell is monitored in real time. When the voltage of an individual battery cell exceeds the overvoltage threshold and the duration exceeds the preset anti-shake delay time, the individual battery cell is determined to have exceeded the limit and marked as a transient overvoltage point.

[0049] Based on the electrical topology connection list of the battery pack, a graph-based adjacency matrix is ​​constructed using a two-dimensional array to represent the electrical topology relationships between individual cells. A breadth-first search algorithm is employed, traversing the adjacency matrix layer by layer from the transient overvoltage point to find individual cells adjacent to the transient overvoltage point in the switch control loop, forming a peripheral candidate set. The calculated and recorded health state factors, temperature deviations, and state of charge values ​​of each individual cell in the peripheral candidate set are retrieved and compared, then sorted. Individual cells in the peripheral candidate set that do not meet the preset safety conditions are eliminated. Among the remaining individual cells, an iterative comparison algorithm is used to find the minimum value, selecting the individual cell with the lowest state of charge as the priority for power reception, preventing the misjudgment of poorly functioning high-temperature or aging individual cells as priority recipients.

[0050] For example, the total current of the main circuit is continuously acquired at a sampling frequency of 100Hz. If the absolute value of the continuously acquired total current of the main circuit is greater than or equal to a preset value (for a 100Ah battery pack, the absolute value of the total current of the main circuit is greater than or equal to 5A), it is determined that the battery is currently in a charging / discharging state. If the absolute value of the total current of the main circuit is lower than the preset value, it is determined that the battery is currently in an idle state. The operating condition compensation parameter corresponding to the charging / discharging state is retrieved. If the battery is currently in a charging state, the corresponding operating condition compensation parameter -0.15V is retrieved.

[0051] To determine transient overvoltage anomalies and filter out artificially high polarization caused by high-rate charging and discharging, a relationship was established between the overvoltage threshold, the absolute overvoltage limit, and the operating condition compensation parameters. The specific relationship is as follows:

[0052] in, Indicates the overvoltage threshold. This represents the absolute overvoltage limit value, which is obtained by reading from the controller's non-volatile memory, pre-programmed into the controller. This represents the operating condition compensation parameter, which is obtained by matching the corresponding constant based on the operating condition type determined by the total current of the main circuit.

[0053] Subtracting the operating condition compensation parameter -0.15V from the absolute overvoltage limit of 4.25V yields an overvoltage threshold of 4.4V. Real-time monitoring of voltage surges in each individual battery cell is performed. When the voltage of a battery cell, for example, logic number 5, reaches 4.42V and exceeds the overvoltage threshold, a timing event is initiated. If the duration of the overvoltage condition exceeds the preset anti-shake delay time of 2 seconds (corresponding to 200 consecutive control cycles), an overvoltage is determined, and logic number 5 is marked as a transient overvoltage point.

[0054] Based on the electrical topology connection list of the battery pack, a predetermined number of adjacent cells in the switch control circuit are located for transient overvoltage points, such as the cell with logic number 15. Four cells with a connection count of two or less are then located, forming a candidate set of cells with logic numbers 13, 14, 16, and 17. The health status factor, temperature deviation, and state of charge (SOC) values ​​of these four cells are read. Cells with a health status factor not exceeding a preset safety threshold (e.g., 0.75) or exceeding the temperature deviation limit are eliminated. For example, cell number 14 is removed because its health status factor is 0.7, and cell number 17 is removed because its current temperature reaches 45°C, exceeding the temperature safety limit. Among the remaining cells that meet the preset safety conditions (cells with logic numbers 13 and 16), they are sorted according to their SOC. Select the cell with the lowest state of charge. Assuming that the state of charge of the cell with logic number 16 is 0.35, set the cell with logic number 16 as the target for relieving transient overvoltage.

[0055] Thus, by dynamically adjusting the overvoltage threshold in conjunction with the operating condition compensation parameters and introducing a time-shaking delay mechanism, the false high voltage judgment caused by transient high-frequency noise or current change is effectively eliminated, and the real transient overvoltage point is accurately identified. Based on the graph theory adjacency matrix, the breadth-first search algorithm combined with the multi-dimensional safety filtering mechanism finds the nearest target that meets the preset safety conditions and has the lowest state of charge, cutting off the hidden danger link of high-temperature single cells or aging single cells being mistakenly put as power receiving objects. This not only quickly eliminates the safety hazard of transient overvoltage at local nodes, but also avoids the high heat loss caused by long-distance cross-module energy transfer.

[0056] S3. Determine the duty cycle and perform balanced control.

[0057] In an optional embodiment, the smaller value of the health status factors of the energy supplier and receiver is extracted and multiplied by a baseline safety constant to obtain the initial limiting current. Combining Ohm's law with the introduced lumped-parameter thermodynamic model, the maximum permissible temperature rise caused by the sum of the ohmic resistances of both the supplier and receiver is transformed into impedance thermal current limiting based on the equivalent heat transfer formula. The minimum value between the initial limiting current and the impedance thermal current limiting is taken as the maximum permissible equilibrium current limit. The real-time voltage difference between the energy supplier and receiver is obtained and divided by the sum of their series ohmic resistances to calculate the expected current. The smaller value between the expected current and the maximum permissible equilibrium current limit is taken as the target current. The actual equilibrium current is collected, and the deviation between the target current and the actual equilibrium current is input into a proportional-integral-derivative (PID) control algorithm as a preset transfer function for joint calculation and dynamic adjustment. The operation outputs a pulse width modulation duty cycle value that fluctuates between 0 and 100%, which is used as the energy transfer duty cycle. The energy transfer duty cycle is converted into a count value of the underlying timer detection and comparison register, and a control signal is output to drive the power switch in the equalization circuit to turn on and off, thereby performing energy transfer.

[0058] In an optional embodiment, the current limit value is calculated based on the ohmic internal resistance and health status factor of both the supply and demand sides, including the following steps: extracting the smaller value of the health status factor of both the supply and demand sides and multiplying it by a reference safety constant to obtain the initial limit current; calculating the impedance thermal current limit according to the relationship; and taking the minimum value between the initial limit current and the impedance thermal current limit as the current limit value.

[0059] To prevent excessive equalization current from causing aging and thermal runaway of the auxiliary battery, a relationship was established between impedance thermal current limiting, maximum allowable temperature rise, and ohmic internal resistance. The specific relationship is as follows:

[0060] in, Indicates impedance thermal current limiting. This represents the maximum permissible temperature rise, obtained by extracting a pre-calibrated thermal stability threshold for the battery material. The ohmic internal resistance of the energy supplier is obtained by measuring using the pulsed current injection method. The ohmic internal resistance of the energy receiver is measured using the pulsed current injection method. This represents the preset equivalent thermal resistance, which is obtained by solidifying parameters through a heat dissipation model between a single cell and the environment.

[0061] In an optional embodiment, the real-time voltage difference and current limit value are input into a proportional-integral-derivative (PID) control algorithm to determine the duty cycle for balancing, including the following steps: dividing the real-time voltage difference by the sum of the ohmic internal resistances of the supply and demand sides to obtain the expected current; taking the smaller value between the expected current and the current limit value as the target current; acquiring the actual balancing current; inputting the deviation between the target current and the actual balancing current into the PID control algorithm to calculate the duty cycle and generate control commands to execute balancing.

[0062] To achieve precise control of the energy transfer process, a proportional-integral-differential relationship was established between the duty cycle control quantity and the current deviation. The specific relationship is as follows:

[0063] in, This indicates the control quantity, i.e., the duty cycle. The deviation is represented by subtracting the target current from the actual equilibrium current. This represents the proportional gain parameter, which is obtained through configuration parameters pre-written into the controller. This represents the integral gain parameter, which is obtained through configuration parameters pre-written into the controller. This represents the differential gain parameter, which is obtained through configuration parameters pre-written into the controller.

[0064] For example, parameters such as the energy supplier's internal resistance of 3mΩ and health status factor of 0.9, and the energy receiver's internal resistance of 4mΩ and health status factor of 0.85 are extracted and substituted into a preset aging constraint function. The smaller value of the health status factors of the supply and demand sides, i.e., 0.85, is extracted and combined with the temperature rise constraint caused by the sum of the internal resistances of 7mΩ, to calculate the maximum allowable equalization current limit value to prevent the aging of the additional battery, which is equal to 2.5A. The real-time voltage difference between the energy supplier and the receiver 0.5ms before the equalization execution closes is collected, which is equal to 180mV. The real-time voltage difference of 180mV is divided by the sum of the internal resistances of the supply and demand sides, 7mΩ, to calculate the expected current. The expected current is compared with the maximum allowable equalization current limit value of 2.5A, and the smaller value of 2.5A is taken as the target current. The actual equalization current is set to 2.0A. The deviation between the target current of 2.5A and the actual equalization current of 2.0A is calculated to be 0.5A. This deviation of 0.5A is input into the proportional-integral-derivative (PID) control algorithm for joint calculation, and the duty cycle percentage is converted to be 45.5%. A pulse width modulation (PWM) control command is generated, outputting a PWM square wave signal with a switching frequency of 100kHz and a duty cycle of 45.5%, and sent to the corresponding switch to perform energy transfer output.

[0065] In this way, by integrating the health status factor and the impedance thermal current limiting caused by the ohmic internal resistance, the maximum allowable equalization current limit value is accurately calculated, which cuts off the risk of irreversible secondary damage and thermal runaway caused by excessive equalization current to fragile single cells. By combining the real-time voltage difference and the preset transfer function to adjust the working duty cycle through proportional-integral-derivative feedback, a stable and efficient energy transfer is achieved, which reduces the additional heat loss in the equalization process while ensuring system safety.

[0066] S4. Multidimensional joint convergence determination terminates equilibrium.

[0067] In an optional embodiment, the standard deviation calculation function from the mathematical operation library is periodically called in the main control loop to calculate the standard deviation of the current state potential index of all individual cells globally. The rate of change is obtained by subtracting the average state potential index of the previous cycle from the average state potential index of the current calculation cycle, and dividing the calculated difference by the sampling time interval. Simultaneously, the discrete variance of the real-time internal equalization current data sequence flowing through the equalization circuit inductor is calculated. If the discrete variance is continuously less than a set variance threshold for a preset sampling period, the internal equalization current is considered stable. When the standard deviation is lower than a first convergence threshold, and the absolute value of the rate of change is lower than a second convergence threshold, and the internal equalization current is considered stable, and the result of the logical AND operation of the three Boolean logic conditions is true, the controller calls the pulse width modulation stop function, cuts off all equalization drive signals, enters sleep mode, and terminates the equalization process.

[0068] In an optional embodiment, the process terminates when the standard deviation and rate of change of the state potential index are lower than a first convergence threshold and a second convergence threshold, respectively, and the equilibrium current is stable. This includes the following steps: calculating the standard deviation of the state potential index of all individual cells; subtracting the average value of the state potential index of the previous cycle from the average value of the current cycle and dividing by the sampling time interval to obtain the rate of change; calculating the discrete variance of the actual equilibrium current data sequence; determining that the equilibrium current is stable when the discrete variance is continuously less than a set variance threshold for a preset sampling period; and cutting off the equilibrium drive signal when the standard deviation is lower than the first convergence threshold, the absolute value of the rate of change is lower than the second convergence threshold, and the equilibrium current is determined to be stable.

[0069] For example, the controller periodically calculates the standard deviation of the state potential index of all individual cells in the main control loop, performs discrete variance calculation on the real-time internal equalization current data sequence flowing through the equalization circuit inductor, and determines that the internal equalization current has reached stability if the discrete variance is less than the set variance threshold for 10 consecutive sampling periods.

[0070] Two lithium iron phosphate battery packs of 100 cells each with the same initial state and manufactured in the same batch were selected and labeled as the experimental group and the control group, respectively, for a comparative experiment. The experimental conditions involved placing both battery packs in a test chamber at an ambient temperature of 40°C and performing dynamic charge-discharge cycle tests. The charge-discharge rate was set to 1C, and the test cycle totaled 500 cycles. The experimental group employed a state potential index assessment, transient overvoltage mitigation, and a duty cycle energy transfer control mechanism based on an aging constraint function to perform equalization. When the standard deviation of the state potential index was lower than the first convergence threshold, and the absolute value of the rate of change was lower than the second convergence threshold, and the internal equalization current was determined to be stable, all equalization drive signals were cut off, terminating energy transfer. The control group employed a traditional passive equalization strategy based on a single voltage threshold for equalization testing.

[0071] like Figure 3 As shown, after 500 charge-discharge cycles, the maximum single-cell voltage difference of the experimental group battery pack decreased from the initial 35mV and stabilized at 15mV at the end of the charge-discharge cycle. The maximum ohmic internal resistance increase of a single cell was only 0.5mΩ, and the usable remaining capacity of the entire battery pack remained at 92.5% of the initial capacity. There were 0 transient overvoltage alarms in the experimental group during the test cycle. In contrast, the maximum single-cell voltage difference of the control group battery pack increased to 55mV, the maximum ohmic internal resistance increase of a single cell reached 1.8mΩ, the usable remaining capacity decreased to 85.2% of the initial capacity, and 12 transient overvoltage alarms occurred during the test cycle. The data comparison shows that the multi-condition Boolean logic joint convergence mechanism comprehensively and objectively determined the global single-cell equilibrium convergence state, cutting off the hidden danger of the equilibrium system falling into a dead loop and repeated oscillations caused by a single threshold trigger, significantly improving the equilibrium efficiency, safety reliability, and overall service life of the battery pack.

[0072] This invention also discloses a battery equalization control system based on dynamic differential pressure, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement a battery equalization control method based on dynamic differential pressure according to the present invention.

[0073] The aforementioned battery equalization control system based on dynamic differential pressure also includes other components well known to those skilled in the art, such as communication buses and communication interfaces. Their settings and functions are known in the art and will not be described in detail here.

Claims

1. A method for balancing control of storage batteries based on dynamic pressure difference, characterized by, include: S1. Collect the voltage, temperature, and charge / discharge current of individual cells in the battery pack, and measure the ohmic internal resistance of the individual cells. Based on the temperature and charge / discharge current, obtain the state of charge, temperature deviation, and health state factor, and construct a state potential index. Classify individual cells with a state potential index greater than a first threshold into the energy supply set, and classify individual cells with a state potential index less than a second threshold among those that meet preset safety conditions into the energy receiving set. The decay rate is calculated using the cumulative number of charge / discharge cycles and the number of reference cycles. The difference between 1 and the decay rate is compared with a preset minimum positive value, and the larger value is taken as the health state factor. S2. Identify individual cells whose voltage exceeds a preset overvoltage threshold as transient overvoltage points. Among the individual cells adjacent to each other in the switch control circuit that meet the preset safety conditions, select the one with the lowest state of charge as the target for facilitating transient overvoltage points. S3. Take the energy supply set and energy receiving set or transient overvoltage point and diversion target as the supply and demand sides, calculate the current limit value based on the ohmic internal resistance and health status factor of the supply and demand sides, and input the real-time voltage difference and current limit value into the proportional integral derivative control algorithm to determine the duty cycle and perform balancing. S4. The process terminates when the standard deviation and rate of change of the state potential index are lower than the first convergence threshold and the second convergence threshold, respectively, and the equilibrium current is stable.

2. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The process of obtaining the state of charge, temperature deviation, and health status factor based on the temperature and the charge / discharge current, and constructing a state potential index, includes: The state of charge is obtained based on the charging and discharging current using the ampere-hour integration method; The deviation between the temperature and the preset standard operating temperature range is taken as the temperature deviation and converted into a temperature suitability deduction factor. The state potential index is obtained by weighting and summing the state of charge, the temperature suitability deduction factor, and the health state factor, wherein the weight coefficient of the temperature suitability deduction factor is negative.

3. The battery equalization control method based on dynamic voltage difference according to claim 1, characterized in that, The preset safety conditions specifically include: the temperature deviation is less than a preset temperature limit value and the health status factor is greater than a preset safety threshold value; Obtaining the first threshold and the second threshold includes: calculating the average baseline value of the state potential index of all individual cells; The first threshold is obtained by adding the product of the absolute value of the average benchmark value and the first floating coefficient to the average benchmark value. The second threshold is obtained by subtracting the absolute value of the average benchmark value from the product of the second floating coefficient.

4. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The calculation of the current limit value based on the ohmic internal resistance and health status factor of both the supply and demand sides includes: The smaller value among the health status factors of the supply and demand sides is extracted and multiplied by the reference safety constant to obtain the initial limiting current; According to the relation: Calculate impedance thermal current limiting ,in For the maximum allowable temperature rise, and These are the ohmic internal resistances of the supply and demand sides, respectively. The preset equivalent thermal resistance; The minimum value between the initial limiting current and the impedance thermal current limit is taken as the current limiting value.

5. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The measurement of the ohmic internal resistance of a single cell includes: A DC test pulse current with a constant amplitude and a pulse width in the millisecond range is injected into the individual battery cell; The terminal voltage of the individual cell is collected before injection as the unperturbed initial voltage; The terminal voltage of the single cell is collected at the transient step moment before polarization occurs during injection as the test final voltage. The ohmic internal resistance is obtained by dividing the absolute value of the difference between the final test voltage and the unperturbed initial voltage by the amplitude of the DC test pulse current.

6. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The single cell whose identification voltage exceeds a preset overvoltage threshold is considered a transient overvoltage point, including: When the voltage of a single battery cell exceeds the overvoltage threshold and the duration exceeds the preset anti-shake delay time, the single battery cell is determined to be over-limit and marked as the transient overvoltage point. The setting of the overvoltage threshold includes: determining whether the operating condition is a charging / discharging state or a static state based on the total current of the main circuit, retrieving the corresponding operating condition compensation parameters, and performing an algebraic sum operation with the absolute overvoltage limit value to obtain the overvoltage threshold.

7. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The step of selecting the cell with the lowest state of charge as the target for mitigation from adjacent cells in the switch control circuit that meet preset safety conditions as the transient overvoltage point includes: Find individual cells adjacent to the transient overvoltage point in the switch control loop to form a surrounding candidate set; Individual cells that do not meet the preset safety conditions are removed from the surrounding candidate set; The cell with the lowest state of charge among the remaining individual cells is selected as the target cell for conduction.

8. The battery equalization control method based on dynamic differential pressure according to claim 1, characterized in that, The step of inputting real-time voltage difference and current limit value into a proportional-integral-derivative control algorithm to determine the duty cycle for balancing includes: The expected current is obtained by dividing the real-time voltage difference by the sum of the ohmic internal resistances of the supply and demand sides. The smaller value between the expected current and the current limit value is taken as the target current; Collect the actual equalization current; The deviation between the target current and the actual equalization current is input into the proportional-integral-derivative control algorithm to calculate the duty cycle and generate control commands to execute the equalization.

9. A battery equalization control method based on dynamic differential pressure according to claim 8, characterized in that, The process terminates when the standard deviation and rate of change of the state potential index are lower than the first convergence threshold and the second convergence threshold, respectively, and the equilibrium current is stable, including: Calculate the standard deviation of the state potential index for all individual cells; The rate of change is obtained by subtracting the average value of the state potential index of the current period from the average value of the previous period and dividing by the sampling time interval. The discrete variance is calculated for the actual equalization current data sequence. When the discrete variance is continuously less than a set variance threshold for a preset sampling period, the equalization current is determined to be stable. When the standard deviation is lower than the first convergence threshold, the absolute value of the rate of change is lower than the second convergence threshold, and the equalization current is determined to be stable, the equalization drive signal is cut off.

10. A battery balancing control system based on dynamic differential pressure, characterized in that, include: A processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement a battery equalization control method based on dynamic differential pressure according to any one of claims 1-9.

Images