A successive approximation equilibrium method
By using a successive approximation equalization method and utilizing voltage and temperature data of the battery pack in a static state, hardware requirements are simplified, and high-precision SOC difference identification and equalization are achieved. This solves the problems of high hardware cost and large error in existing technologies, and improves the equalization efficiency and lifespan of the battery pack.
Patent Information
- Application Number
- CN202511352604.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-22
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2045-09-22
AI Technical Summary
Existing battery pack equalization methods are highly dependent on measurement data, have high hardware costs and system complexity, and are limited by error and algorithm complexity, making it difficult to achieve high-precision SOC difference identification and effective equalization.
A successive approximation equalization method is adopted. By monitoring the static state of the battery pack and using voltage and temperature data, the SOC difference is adjusted successively and approximatingly. This simplifies the process to rely solely on voltage and temperature monitoring, avoiding complex algorithms and reducing hardware requirements.
It achieves high-precision SOC difference identification and equalization, reduces hardware costs and system complexity, avoids the "equalization tidal" phenomenon, improves equalization efficiency and battery life, and supports battery health status assessment.
Smart Images

Figure QLYQS_1 
Figure QLYQS_2 
Figure QLYQS_3
Abstract
Description
Technical Field
[0001] This invention relates to the field of equalization methods for series battery packs, and more particularly to a successive approximation equalization method. Background Technology
[0002] In modern electrochemical energy storage systems, especially in battery management systems (BMS) for new energy vehicles, maintaining precise state of charge (SOC) consistency among the series-connected cells within a battery pack is crucial for ensuring system performance, lifespan, and safety. Equalization technology, as the core function for achieving this goal, can generally be broken down into three main stages: measurement (obtaining differences in cell SOC), decision-making (developing equalization strategies based on these differences and system conditions), and execution (driving the hardware to complete energy transfer). Among these, the measurement stage is fundamental and the most challenging, directly determining the accuracy of subsequent decisions.
[0003] Current mainstream equalization technologies suffer from the following drawbacks: excessive reliance on and complexity of measurement data. Many existing equalization algorithms, especially those requiring high-precision SOC difference identification, often need to collect physical quantities other than individual cell voltages, such as charging and discharging currents. This not only increases hardware costs (requiring high-precision current sensors) and system complexity but may also introduce additional power consumption and measurement errors. Accurate current measurement is even more difficult under low current or quiescent conditions. Even equalization methods based on open-circuit voltage (OCV) rely on high-precision OCV-SOC curves and have extremely high requirements for the battery pack's quiescent time, which is difficult to meet in practical applications.
[0004] Algorithm complexity and potential errors: To achieve accurate SOC difference identification and equalization decisions with limited measurement data, some algorithms employ complex mathematical models or machine learning methods. This significantly increases the computational demands on the processor, limiting their application in resource-constrained BMS systems. Furthermore, even with complex algorithms, "equalization tidal" phenomena can still occur due to model simplification, parameter drift, or accumulated computational errors. "Equalization tidal" refers to the system performing reverse operations to correct over-equalization errors after over-equalization, resulting in additional charging and discharging, reducing equalization efficiency and battery life.
[0005] Therefore, there is an urgent need for a technology that can simplify measurement requirements (e.g., rely solely on voltage), reduce algorithm complexity, and at the same time achieve high-precision SOC difference identification and effective equalization to overcome the limitations of existing solutions. Summary of the Invention
[0006] The present invention provides a successive approximation-type equilibrium method, which solves the technical problem of the complexity of existing equilibrium methods.
[0007] Some implementation schemes for solving the above-mentioned technical problems include: A successive approximation equilibrium method includes the following steps: Step 10, initialize a "single unit capacity array" using the rated capacity of each unit, denoted as "". " Indicates the cell number; Step 20: Continuously monitor the cell voltage and identify the event "the battery pack is in a quiescent state". Once the condition is confirmed, proceed to step 30; otherwise, stay in step 20 and loop. Step 30: Measure the voltage and temperature of each individual cell. Based on the OCV-T-SOC data provided by the manufacturer, find the current SOC value of each cell and construct the "original SOC array," denoted as "". Step 40, according to and Calculate and construct the "balanced execution power" and denot it as " "" and "Difference in target SOC after execution" are denoted as " "Two arrays; Step 50, according to..." Values are balanced; step 60, after balancing, values are balanced according to... After the actual power transfer is completed, the individual cell voltage is continuously monitored and the event "the battery pack is in a quiescent state" is identified. Step 70: Collect voltage and temperature data and find the current SOC value of each cell. Measure and look up the SOC of each individual cell to construct a "SOC difference array after equalization," denoted as "...". Step 80, according to and [i] Calculate the "real-time unit capacity", denoted as " ”; Step 90, using Value substitution The value in, using The value in [i] is replaced The value in the table is returned to step 40. Preferably, the method for identifying whether the battery pack is in a quiescent state includes the following steps: Step 1, measure the voltage of each cell and record it as follows: Step 2, measure and record the voltage of each individual cell. Calculate the voltage difference between two adjacent voltages. ,Will The value assigned to Step 3, Compare Are all less than If so, record the current time as Skip to step 4, otherwise skip to step 2; Step 4, measure the individual cell voltage and record it as... Calculate the voltage difference between two adjacent voltages. ,Will The value assigned to Step 5, check Are all less than If yes, skip to step 6; otherwise, skip to step 1. Step 6: Calculate the current time. right jet lag ,Compare Does it exceed If the value exceeds the limit, proceed to step 7; otherwise, proceed to step 4. Step 7: Calculate the voltage difference between the two ends. ,Compare Are all less than If not, skip to step 1; if yes, it indicates the battery pack meets the resting conditions. Preferably, the method for obtaining the equalization charge includes: after confirming the battery pack has entered a resting state, measuring the individual cell voltage and then finding the SOC value of the individual cell according to the OCV-T-SOC table provided by the manufacturer; based on these SOC values, determining the individual cell capacity. And calculate the specific amount of electricity that needs to be balanced based on different balancing circuits. The method is as follows: Measure the individual cell voltage, consult the OCV-T-SOC table provided by the manufacturer, and construct a "current SOC array" denoted as "". The "desired transfer SOC array" is calculated for different equilibrium methods and denoted as "". ”; If it is a discharge-only passive equalization, take The minimum value in is denoted as " "; If it is a rechargeable type (power comes from an external power source rather than a battery pack) active balancing, take..." The maximum value in is denoted as " "; If it is a charge / discharge type active balancing, calculate..." The arithmetic mean is denoted as " ";calculate ; according to , Calculate the equilibrium charge using the k value: k,k A decimal number whose value is between 0 and 1; according to The "ideal SOC difference array" after balancing with k value is denoted as , (1-k) Preferably, the method for real-time cell capacity measurement includes: Measure the individual cell voltage, consult the OCV-T-SOC table provided by the manufacturer, and construct a "current SOC array" denoted as "". ";Bundle 0 Identify the individual cells and calculate the array of "self-discharge leading to SOC error" for these channels, denoted as... OC [j],j It is all 0 The set composed of the subscripts of the individual units is calculated as follows: Then calculate The median is used as the "self-discharge rate benchmark" and is denoted as " For different equilibrium methods, the "post-transfer SOC difference array" is calculated and denoted as "". ”; If it is a discharge-only passive equalization, take The minimum value in is denoted as " ”; If it is a charging-only active balancing system, take... The maximum value in is denoted as " ”; If it is a charge / discharge type active balancing, calculate... The arithmetic mean is denoted as " ";calculate ; according to , and The estimated "true capacity of a single unit" is denoted as ; If the single-unit label i corresponds to If the value is 0, then continue using the corresponding label. The value as The value; if the single unit label i corresponds to Not equal to 0 and corresponding label and If two values have opposite signs, then continue using the corresponding label. The value as The value; If the single-unit label i corresponds to Not equal to 0 and corresponding label and If two values have the same sign, then use / Compared with the prior art, the present invention has the following advantages: This invention achieves high-precision battery balancing without the need for additional current sensors, relying solely on the battery pack's existing voltage and temperature monitoring capabilities. This significantly simplifies the BMS hardware configuration, reducing material costs and system integration complexity. For devices requiring balancing functionality, this translates to lower manufacturing costs and greater market competitiveness.
[0008] Compared to traditional techniques that require complex mathematical models or machine learning algorithms, this equalization algorithm avoids the computational bottleneck caused by algorithm complexity through a "successive approximation" strategy (the algorithm itself can even run on a regular 8-bit microcontroller), making it more widely applicable in resource-constrained BMS. At the same time, the "successive approximation" characteristic effectively avoids the "equalization tidal" phenomenon, reducing unnecessary energy loss and potential damage to the battery, and improving equalization efficiency and battery life.
[0009] This invention does not rely directly on the actual capacity of the battery after aging, but eliminates the SOC difference through a successive approximation method. This makes it more adaptable to individual differences and aging states of batteries. More importantly, after equalization, this invention can further calculate the capacity differences between batteries, providing valuable data support for battery health assessment and lifespan prediction, and further improving the intelligence level of the battery management system. Detailed Implementation
[0010] The specific embodiments shown below are intended to describe various configurations of the subject matter of the invention and are not intended to represent the only configuration in which the subject matter of the invention can be practiced. The specific embodiments include particular details intended to provide a thorough understanding of the subject matter of the invention. However, it will be clear and apparent to those skilled in the art that the subject matter of the invention is not limited to the specific details shown herein and can be practiced without these specific details.
[0011] Understandably, in this document, relational terms such as “first” and “second” are intended to distinguish one entity or operation from another, and are not intended to expressly or imply any actual relationship or order between these entities or operations.
[0012] The terms “comprising,” “including,” or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase “comprising one…” does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0013] A successive approximation equalization method, characterized by the following steps: Step 10, initializing a "cell capacity array" denoted as "" using the rated capacity of a single cell. " Indicates the unit number; Step 20: Continuously monitor the voltage of individual cells and identify the event that "the battery pack is in a quiescent state". Once the condition is confirmed, proceed to step 30; otherwise, stay in step 20 and loop. Step 30: Measure the voltage and temperature of each individual cell. Based on the OCV-T-SOC data provided by the manufacturer, find the current SOC value of each cell and construct the "original SOC array," denoted as "". Step 40, according to and Calculate and construct the "balanced execution power" and denot it as " "" and "Difference in target SOC after execution" are denoted as " "Two arrays; Step 50, according to..." Values are balanced; Step 60, after balancing is complete, proceed according to... After the actual power transfer is completed, the individual cell voltage is continuously monitored and the event "the battery pack is in a quiescent state" is identified; in step 70, the voltage and temperature are collected and the current SOC value of each cell is found. The SOC of each individual cell is measured and obtained by looking up a table, and a "SOC difference array after equalization" is constructed, denoted as " Step 80, according to and [i] Calculate the "real-time unit capacity", denoted as " ”; Step 90, using Value substitution The value in, using The value in [i] is replaced Return to step 40 if the value in the table is correct. Method to confirm the battery pack is in a quiescent state: The battery pack is accurately determined to be in a static state by voltage (and only by voltage). Only when the battery pack is in a static state can the accurate SOC value of each cell be found through VOC-T-SOC data.
[0014] Based on the general characteristics of electrochemical secondary batteries (including but not limited to lithium batteries, sodium batteries, lead-acid batteries, etc., hereinafter collectively referred to as "secondary batteries"), a significant (detectable) voltage drop occurs at the moment charging ends. This phenomenon can be identified by taking the difference between two consecutive voltage measurements (current voltage vs. previous voltage) and comparing it to a threshold value (this threshold value is related to the battery's chemical system, capacity configuration, and charger type; it can be obtained through type testing during the design phase; this threshold value is denoted as "the threshold value"). Identification is performed by comparison. This is based on the fact that the secondary battery has been left undisturbed (neither charging nor discharging) for a sufficiently long period (this time will be referred to below as ""). The individual cell voltage tends to stabilize, and the definition of stabilizing is that the individual cell voltage... The difference between the maximum and minimum values measured continuously within a time period is a threshold (this threshold is related to the battery's chemistry; the manufacturer can provide this value, which is denoted as "..."). During the resting process, as long as the resting conditions are disrupted (the battery is charged and discharged), the change in individual cell voltage is significant (detectable). The detection method is to calculate the difference between two consecutive voltage measurements (current voltage vs. previous voltage) and compare it with a threshold value (this threshold value is related to the battery's chemical system and can be provided by the manufacturer; this threshold value is denoted as ""). Identification is performed by comparing the absolute values of ”.
[0015] The specific testing method for the static state is as follows: Step 1, measure the voltage of each cell and record it as follows: Step 2, measure and record the voltage of each individual cell. Calculate the voltage difference between two adjacent voltages. ,Will The value assigned to Step 3, Compare Are all less than If so, record the current time as Skip to step 4, otherwise skip to step 2; Step 4, measure and record the voltage of each individual cell. Calculate the voltage difference between two adjacent voltages. ,Will The value assigned to Step 5, check Are all less than If yes, skip to step 6; otherwise, skip to step 1. Step 6, calculate the current time right jet lag ,Compare Does it exceed If the value exceeds the limit, proceed to step 7; otherwise, proceed to step 4. Step 7: Calculate the voltage difference across the two terminals. ,Compare Are all less than If not, skip to step 1; if yes, it means the battery pack meets the resting conditions.
[0016] The method for calculating the difference between the target SOC and the battery level after equalization is as follows: After confirming that the battery pack has entered a static state, measure the individual cell voltage and then find the SOC value of each cell according to the OCV-T-SOC table provided by the manufacturer. Then, based on these SOC values, calculate the individual cell capacity. And calculate the specific amount of electricity that needs to be balanced (transferred) based on different balancing circuits. The method is as follows: Measure the individual cell voltage, consult the OCV-T-SOC table provided by the manufacturer, and construct a "current SOC array" denoted as "". ”; The "desired transfer SOC array" is calculated for different equilibrium methods and denoted as "". ”; If it is a discharge-only passive equalization, take The minimum value in is denoted as " ”; If it is a rechargeable type (power comes from an external power source rather than a battery pack) active balancing, take... The maximum value in is denoted as " "; If it is a charge / discharge type active balancing, calculate..." The arithmetic mean is denoted as " ”; calculate ; according to , Calculate the equilibrium charge using the k value: k, Here, k is a key value. k The value must be a decimal between 0 and 1, typically between 0.5 and 0.75; it is crucial for the "successive approximation" process. Assume... And if all measurements and calculations are accurate, then It means accurate, in other words, strictly according to Performing a balancing process could eliminate the SOC differences between monomers in one go, but this is clearly impractical due to measurement and calculation errors. The value will inevitably contain errors, especially since the method of this technology was initially initialized according to the design rated capacity. However, due to manufacturing errors, aging, and other reasons, the actual capacity of the individual cells is different. The initial values must be different. Errors can lead to over-balancing, which in turn causes imbalance tidal currents. Increasing this k coefficient can eliminate the possibility of over-balancing, because the existence of k is equivalent to "forcing" the balanced transfer power to be lower than the calculated power, in other words, forcing "under-balancing". The specific parameter of this value is determined by the designer's assessment of measurement and calculation errors, but it must be between 0 and 1. The range of k values is specific to this technology, and the specific value of k is not considered a technical characteristic. Calculated... This refers to the equalization of energy usage, measured in Ah. A positive number indicates the amount of electricity (in Ah) needed to be charged, while a negative number indicates the amount of Ah needed to be released. According to... The "ideal SOC difference array" after balancing with k value is denoted as , (1-k)Methods for updating real-time unit capacity: Assuming there are no measurement or calculation errors, the SOC difference between monomers will decrease as expected after equalization. Even within the same system, if the measurement and calculation errors of each individual unit are consistent, the consistency error is eliminated after subtraction, and the final SOC difference is equal to... Therefore, once equilibrium is reached, the SOC difference is not equal to... There can only be two main reasons: 1) Self-discharge of individual cells causing deviation; 2) Initial capacity of individual cells... The preset values are inconsistent. This is the primary cause, and can be eliminated using the following method: Measure the individual cell voltage, consult the manufacturer's OCV-T-SOC table, and construct a "current SOC array" denoted as "". ”; First of all 0 Identify the individual cells and calculate the array of "self-discharge leading to SOC error" for these channels, denoted as... OC[j],j It is all 0 The set composed of the subscripts of the individual units is calculated as follows: Then calculate The median is used as the "self-discharge rate benchmark" and is denoted as " For different equilibrium methods, the "post-transfer SOC difference array" is calculated and denoted as "". ”; If it is a discharge-only passive equalization, take The minimum value in is denoted as " ”; If it is a rechargeable type (power comes from an external power source rather than a battery pack) active balancing, take... The maximum value in is denoted as " "; If it is a charge / discharge type active balancing, calculate..." The arithmetic mean is denoted as " ”; calculate ; according to , and The estimated "true capacity of a single unit" is denoted as If the single-unit label i corresponds to If the value is 0, then continue using the corresponding label. The value as The value; If the single-unit label i corresponds to Not equal to 0 and corresponding label and If two values have opposite signs, then continue using the corresponding label. The value as The value; If the single-unit label i corresponds to Not equal to 0 and corresponding label and If two values have the same sign, then use / .
[0017] The technical solution of the present invention and its corresponding details have been described above. It is understood that the above description is only some implementation schemes of the technical solution of the present invention, and some details may be omitted in the specific implementation.
[0018] Furthermore, in some embodiments of the above invention, multiple embodiments may be combined; however, due to space limitations, all such combinations will not be listed here. Those skilled in the art can freely combine and implement the above embodiments according to their needs to obtain a better application experience.
Claims
1. A successive approximation equilibrium method, characterized in that, The process includes the following steps: Step 10, initialize a "single-unit capacity array" using the rated capacity of each individual unit. This "single-unit capacity array" is denoted as "". " Indicates the cell number; Step 20, continuously monitors the cell voltage and identifies the event "the battery pack is in a quiescent state". Once the event is confirmed, proceed to step 30; otherwise, loop through step 20; Step 30, measures the voltage and temperature of each cell, finds the current SOC value of each cell based on the OCV-T-SOC data provided by the manufacturer, and constructs the "original SOC array" denoted as "". Step 40, according to and Calculate and construct the "balanced execution power" denoted as " "Difference between target SOC and "after execution" is denoted as " "Two arrays; Step 50, according to..." Values are balanced; step 60, after balancing, values are balanced according to... After the actual power transfer is completed, the individual cell voltage is continuously monitored and the event "the battery pack is now in a quiescent state" is identified; in step 70, voltage and temperature are collected, and the current SOC value of each cell is found. The SOC of each individual cell is measured and obtained by looking up a table, and a "SOC difference array after equalization" is constructed, denoted as "...". Step 80, according to and Calculate the "real-time single-unit capacity", denoted as " Step 90, using Value substitution The value in, using Value substitution Return to step 40 with the value in the table; The method for obtaining the equalization power includes: after confirming that the battery pack has entered a static state, measuring the individual cell voltage and then looking up the SOC value of the individual cell according to the OCV-T-SOC table provided by the manufacturer; based on these SOC values, determining the individual cell capacity. And calculate the specific amount of electricity that needs to be balanced based on different balancing circuits. The method is as follows: Measure the individual cell voltage, consult the OCV-T-SOC table provided by the manufacturer, and construct a "current SOC array" denoted as "". "; The "desired transfer SOC array" is calculated for different equilibrium methods and denoted as " "; If it is only a discharge-type passive equalization, take The minimum value in is denoted as " "; If it is a charging-only active balancing type, take" The maximum value in is denoted as " "; If it is a charge / discharge type active balancing, calculate..." The arithmetic mean is denoted as " ";calculate ;according to , and Value calculation for balanced power consumption: , A decimal number whose value is between 0 and 1; according to and The "ideal SOC difference array" after value calculation and equilibrium is denoted as , The rechargeable type refers to a type where the power source is an external power supply rather than a battery pack.
2. The successive approximation equalization method according to claim 1, characterized in that: The method for identifying whether a battery pack is in a quiescent state includes the following steps: Step 1, measure the voltage of each individual cell and record it as follows. Step 2, measure and record the voltage of each individual cell. Calculate the voltage difference between two adjacent voltages. ,Will Assign to Step 3, Compare Are all less than If so, record the current time as Skip to step 4, otherwise skip to step 2; Step 4, measure the individual cell voltage and record it as... Calculate the voltage difference between two adjacent voltages. ,Will Assign to Step 5, check Are all less than If yes, skip to step 6; otherwise, skip to step 1. Step 6: Calculate the current time. right jet lag ,Compare Does it exceed If the value exceeds the limit, proceed to step 7; otherwise, proceed to step 4. Step 7: Calculate the voltage difference between the two ends. ,Compare Are all less than If not, skip to step 1; if yes, it means the battery pack meets the resting conditions.
3. The successive approximation equalization method according to claim 1, characterized in that: Methods for real-time cell capacity measurement include: measuring cell voltage, consulting the manufacturer's OCV-T-SOC table, and constructing a "current SOC array" denoted as "Current SOC". ";Bundle Identify the individual cells and calculate the array of "self-discharge leading to SOC error" for these channels, denoted as... , It is all The set consisting of the subscripts of the individual units is calculated using the following method: Then calculate The median, denoted as the "self-discharge rate benchmark", is used as the "self-discharge rate benchmark". For different equilibrium methods, the "post-transfer SOC difference array" is calculated and denoted as "". ”; If it is a discharge-only passive equalization, take The minimum value in is denoted as " "; If it is a charging-only active balancing type, take" The maximum value in is denoted as " "; If it is a charge / discharge type active balancing, calculate..." The arithmetic mean is denoted as " ",calculate ;according to , and The estimated "true capacity of a single unit" is denoted as If the single-unit label i corresponds to If the value is 0, then continue using the corresponding label. The value as The value; if the single unit label i corresponds to Not equal to 0 and corresponding label and If two values have opposite signs, then continue using the corresponding label. The value as The value; if the single unit label i corresponds to Not equal to 0 and corresponding label and If two values have the same sign, then use .
Citation Information
Patent Citations
Battery pack active equalization system and method
CN114421567A
Efficient equilibrium calculation method based on lithium iron phosphate cell unbalance model
CN117172028A
