Lithium iron phosphate battery consistency calculation method and system
By calculating the moment when the voltage change rate is the largest during the charging process of lithium iron phosphate batteries and integrating the charging capacity, the problem of inaccurate battery consistency evaluation and narrow application scope in the prior art is solved, and a more accurate battery consistency evaluation and a wider range of application are achieved.
Patent Information
- Application Number
- CN202510103061.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-22
- Publication Date
- 2025-05-06
AI Technical Summary
It is difficult for the prior art to accurately evaluate the consistency of lithium iron phosphate batteries, especially in the range of 35%-60% and 70%-95% of SOC, the voltage remains basically unchanged, resulting in large errors in the evaluation of SOC and narrow application range.
The consistency of a single battery or battery system is evaluated by calculating the moment when the voltage change rate is the largest during the battery charging process and performing charging capacity integration within this time range.
This method can more accurately evaluate battery consistency, expand the scope of application, and reduce errors in SOC evaluation.
Smart Images
Figure CN119936670A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to a lithium iron phosphate battery consistency calculation method and system, belonging to the field of lithium iron phosphate battery evaluation. Background Art
[0002] The cruising range is a core indicator that customers pay attention to when purchasing electric vehicles. The consistency of the power system is the key factor that determines the cruising range. Accurately judging the consistency means accurately evaluating the cruising range of the power system.
[0003] The static voltage (OCV) of a battery refers to the stable voltage of the battery after the current is zero for a period of time. Through experiments, the relationship between the static voltage (OCV) of a battery and the battery core charge number (SOC) can be obtained, see Figure 1 . It can be seen that the SOC-OCV curve of lithium iron phosphate battery has two platforms. In the platform area, as the SOC increases, the voltage remains almost unchanged. This leads to a many-to-one mapping between OCV and SOC, making it impossible to judge SOC by OCV. The part where SOC can be calculated by OCV is limited to the stage where the battery SOC is low, and the closer to the first platform, the greater the error.
[0004] The existing technology obtains the SOC of the battery by identifying the static voltage of the battery and finding the corresponding OCV-SOC curve. The SOC-OCV curve of the lithium iron phosphate battery has two platforms at 35%-60% and 70%-95% (schematically illustrated, the specific values of different materials are different), and the voltage is basically unchanged within this range. In the range of 25%-35%, the voltage does not change much with the increase of SOC. This leads to a slight difference in the battery voltage itself, which may correspond to a wide range of SOC, and the evaluation error of battery consistency is large. Only in the range of 0%-25% do the voltage and SOC have a corresponding relationship with a small error, but the application range is narrow. Summary of the invention
[0005] The purpose of the present invention is to provide a lithium iron phosphate battery consistency calculation method and system to solve the problem of inaccurate battery consistency evaluation and narrow application scope.
[0006] To achieve the above object, the solution of the present invention includes: The technical solution of a method for calculating consistency of lithium iron phosphate batteries of the present invention comprises: charging, between two plateau periods when the voltage of the target single cell changes with time, calculating the time interval between the moment when the single cell that earliest reaches the maximum voltage change rate reaches the maximum corresponding voltage change rate and the moment when the single cell that latest reaches the maximum voltage change rate reaches the maximum corresponding voltage change rate among the target single cells; integrating the charging capacity of the target single cells within the time interval; the smaller the integral result, the better the consistency between the target single cells.
[0007] Furthermore, the charging is constant current or constant power.
[0008] Furthermore, the moment when the voltage change rate is the largest is determined by the following steps: the slope of the curve corresponding to the voltage change over time is calculated, and the time period before and after the peak value whose slope is not empty at several moments before and after is found as the screening range of the target point. In the screening range, the moment when the voltage change rate is the largest is taken as the moment when the voltage change rate is the largest.
[0009] Furthermore, if multiple target points satisfy the condition that the slope is greater than the slope within the left and right set ranges, the average of the maximum moment and the minimum moment among these target points is taken as the moment when the voltage change rate is the largest.
[0010] The technical solution of a method for calculating the consistency of a lithium iron phosphate battery of the present invention comprises: charging, between two plateau periods when the voltage of a single cell changes with time, calculating the time interval between the moment when the single cell that earliest reaches its own voltage change rate reaches the maximum corresponding voltage change rate and the moment when the single cell that latest reaches its own voltage change rate reaches the maximum corresponding voltage change rate among all the single cells in the battery system; integrating the charging capacity of the battery system within the time interval; the smaller the integral result, the better the consistency of the battery system.
[0011] Furthermore, the charging is constant current or constant power.
[0012] Furthermore, the moment when the voltage change rate is the largest is determined by the following steps: the slope of the curve corresponding to the voltage change over time is calculated, and the time period before and after the peak value whose slope is not empty at several moments before and after is found as the screening range of the target point. In the screening range, the moment when the voltage change rate is the largest is taken as the moment when the voltage change rate is the largest.
[0013] Furthermore, if multiple target points satisfy the condition that the slope is greater than the slope within the left and right set ranges, the average of the maximum moment and the minimum moment among these target points is taken as the moment when the voltage change rate is the largest.
[0014] A technical solution of a lithium iron phosphate battery consistency calculation system of the present invention includes a processor, and the processor executes a computer program to implement the lithium iron phosphate battery consistency calculation method as described above.
[0015] The beneficial effect of the present invention is: using the narrow window between the first platform and the second platform in the process of battery voltage changing over time, determining the time when each single battery passes through this window and the voltage change rate reaches the peak, integrating the battery charged capacity between the earliest time reaching this moment and the latest time reaching this moment, so as to evaluate the consistency between single batteries. This method is more accurate in evaluating consistency and has a wide range of battery applications. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 It is a SOV-OCV relationship diagram in the prior art; Figure 2 is a voltage curve diagram provided by the present invention; Figure 3 It is an introduction diagram of the slope variation curve diagram of the automatic peak finding method provided by the present invention; Figure 4 It is a curve diagram of slope variation over time provided by the present invention. DETAILED DESCRIPTION
[0017] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention is described in detail in a clear and complete manner in conjunction with the accompanying drawings and embodiments.
[0018] like Figure 2 , similar to the SOC-OCV curve, the stable (constant current or constant power) battery charging curve also has two platforms. There is a short period of time between the two platforms and the voltage rises faster.
[0019] A method for calculating the consistency of a lithium iron phosphate battery of the present invention takes into account that in the voltage-time curve of the lithium iron phosphate battery during the charging process, the speed change between the first platform and the second platform is significantly improved compared with the platform period, and the time range is narrow. Similar to the principle of conveniently comparing the consistency differences between batteries at the time of full charge or emptying, this method identifies a fixed moment in the middle of the use of the lithium iron phosphate battery to compare all single cells. The essence of the generation of this moment is caused by the material structure of the lithium iron phosphate itself, so it can be used as an objective standard. The method using the SOC-OCV mapping before the first platform is not a point in the part close to the first platform, but a wide range, which causes the inherent error of the method itself to be much larger than the method of this patent.
[0020] From the perspective of usage scenario coverage, many vehicles will not be used below 25%, or even if it is below 25%, it does not meet the static voltage value conditions. The advantage of this method is that it only relies on the battery to meet a stable charging process and cross the SOC of about 65%, which can solve the problem that some vehicles cannot be evaluated for consistency due to incomplete charging and no low end.
[0021] Method Example 1: Step 1: Obtain the vehicle's primary charging data. Specifically, the vehicle's primary constant current or constant power charging data is extracted through the big data cloud engine Spare. The data includes time, current, SOC, single cell voltage, etc.
[0022] Step 2: decompress the data extracted in step 1 to make the data into a data format that is easy to process. Specifically, the cell voltage is often compressed into a long field to save storage space. Table 1 below is an example table of data before decompression.
[0023] Table 1 Use the "posexplode(split voltage list, '_')" method in Sparksql to unpack the voltage list to process each cell separately. The Split function separates the data into columns according to a symbol, and the posexplode function unpacks the data with a position, which converts the data into columns and increases the position (because the serial number starts from 0, in order to correspond to the battery serial number, all serial numbers are increased by 1 at the same time) Table 2 is an example of data after decomposition.
[0024] Table 2 According to the data after decomposition, the curve of the voltage change of each single cell over time can be obtained, such as Figure 1 .
[0025] Step 3: Calculate the rate of voltage change. Specifically, calculate the slope of the voltage versus time curve in step 2. Figure 2 As shown, it can be seen that during the actual charging process of the battery, the voltage value will remain unchanged for a period of time. During this period of time, the voltage change rate cannot be simply calculated by direct derivation. At this time, as an optional implementation method, the following method is used for processing: First, we need to use the average of the earliest and latest times of the same voltage value of each cell in each vehicle to represent the time corresponding to this voltage. After processing, a voltage value corresponds to only one time point, and this time is defined as the midpoint of time (distinguished from the original time), and then the voltage change rate is calculated. Then the two values of voltage and time midpoint are checked for monotonicity to ensure that the voltage rises with the increase of time. As long as we intercept the data of a constant current or constant power charging process of a vehicle in step one, all cells will satisfy monotonic increase if there is no data transmission error.
[0026] Step 4: Determine the time corresponding to the fastest voltage change rate.
[0027] The entire voltage-time curve corresponds to five processes in a complete case: a) When charging starts, the battery is polarized and the voltage changes fastest at the initial point, then gradually slows down.
[0028] b) Polarization tends to be stable, and the battery enters the first platform, corresponding to a SOC of about 35%-60%, and the voltage changes very slowly.
[0029] c) The battery enters the second platform from the first platform, corresponding to a SOC of about 60%-70%. The voltage changes gradually faster, reaches the peak and then becomes slower.
[0030] d) The battery enters the second platform, corresponding to SOC of about 70%-95%, and the battery changes very slowly.
[0031] e) The battery enters the final stage of charging, corresponding to SOC greater than 95%.
[0032] Depending on the size of the system consistency difference, some monomers may stop at stage d). At this time, according to the automatic peak search algorithm below, it still does not affect the correct identification of the peak corresponding time.
[0033] However, if the consistency difference increases further, causing the monomer to remain in stage c), it will lead to the problem of unrecognition. This is because there is no target peak without crossing stage c. This will affect quantitative judgment, but not qualitative judgment. This is a situation where the system consistency difference is extremely large.
[0034] The ideal method is to find the peak (extreme) position by taking the derivative again. However, it can be seen that the monotonically increasing voltage curve achieved by adjustment in the previous step has its rough precision feature reappearing on the curve after the slope is calculated. In order to make the calculation result more accurate, the following method (i.e., automatic peak finding method) is used as an implementable method to determine the time: Calculate the slope of the voltage corresponding to the 12 time points before and after each time point.
[0035] The requirements for time points are: 1: The 12 points before and after cannot be empty. This condition determines the screening range of the target point.
[0036] 2: The slope of the target point is greater than or equal to the slopes of points 1-6 adjacent to the left, and greater than the slopes of points 7-12 adjacent to the right.
[0037] The above algorithm with specific numerical values as the limitation is easier to implement through Sparksql, but a more thorough method can be to continuously increase the number of recognition points forward and backward until the difference between the maximum time and the minimum time in the set of time points that meet the conditions is small enough, and the entire time is enough for the little finger to charge the battery by about 5%, which can be implemented through Pyspark.
[0038] The two requirements of step 4 are explained in detail through the following two examples: Explanation of requirement 1: See Figure 3 , the machine recognition logic can be briefly described as a point can be defined as a peak value as long as it is larger (or equal) than the points within a certain range to its left and right. From the perspective of manual recognition, the vertex of the left triangle is not our target. During manual recognition, the brain takes all points of the entire image into consideration, compares the widths of the corresponding peaks of the left and right triangles, and selects the right triangle with a larger width.
[0039] By counting the slope variation curves of a large number of batteries over time, we set the width to 12, that is, if the six points on both sides of a point are less than or equal to this point, then this point is defined as the peak value to exclude the triangle on the left.
[0040] Obviously, the determined value has limitations, but under the combined influence of several conditions where the battery material is limited to lithium iron phosphate, the voltage sampling accuracy is mV, the peak value is within the narrow SOC60-70%, and the actual month only corresponds to 1 to 2% SOC range, this fixed value can meet the needs of use. First of all, the voltage sampling accuracy is mV, and the voltage needs to change by at least 1mV to be calculated and only calculated once, which results in the number of data points being independent of the sampling frequency. Because the upper and lower limits of the charging voltage are relatively fixed, the order of magnitude of the points on the slope curve is fixed rather than arbitrary. In theory, 25 points must be met for calculation, which means that the voltage range during charging must be at least greater than 25mV. In fact, this condition is not difficult to meet.
[0041] A more common method is to continuously expand the left and right boundaries and take the last one as the basis.
[0042] Explanation of requirement 2: See Figure 4, the point (1000, 0.04) is extremely likely to be misrecognized. First, we use the round() function to adjust the precision of the slope data to less than 3 digits after the decimal point, so that some similar slope values are regarded as the same. We observe that due to the limited actual precision level, the target time may not be a single point, but several (e.g., Figure 3 , there are 4), so we allow the points adjacent to the target point to be equal. Therefore, for the left and right adjacent 1 - 6 digits, it is required that the slope is greater than or equal to the adjacent points. However, for a point like (1000, 0.04), for the 7 - 12 digits, it is strictly required to be greater and not equal, aiming to exclude this situation. This assumption itself is quite reasonable because the target position reaches the peak within a narrow SOC range, so it is impossible to have numerous equal points.
[0043] That is, based on the characteristic that the charging slope curve of a single battery is a W - shaped graph, requirement 1 first helps us reduce the peaks at both ends, making the peak range concentrated in the middle part of the W. In the remaining part, the algorithm tries to find a graph similar to the Chinese character '几'. Even if the peaks at both ends are not fully reduced, the remaining parts at both ends will not meet requirement 2 and will not be misrecognized.
[0044] Step Five, calculate the consistency difference between single batteries.
[0045] When calculating the consistency among several single batteries (i.e., taking these several single batteries as target single batteries), for the charging data with constant current or constant power during the charging process, between the two plateau periods of the voltage change over time of the target single battery, after each single battery is processed by step four, there is a unique moment with the maximum voltage change rate, that is, the peak moment. Integrate the charging capacity between the earliest and the latest moments among the target single batteries up to this moment. The smaller the integration result, the better the consistency of the target single batteries.
[0046] Integrate the charging capacity between the peak times of two adjacent single batteries to obtain the consistency difference between the two adjacent single batteries. The smaller the integration result, the better the consistency.
[0047] Method Embodiment 2: Based on Method Embodiment 1, when performing consistency analysis on a lithium iron phosphate battery system, all single batteries in the battery system are taken as target single batteries. Between the two plateau periods of the voltage change over time of all single batteries in the battery system (i.e., target single batteries), after each single battery is processed by step four, there is a unique moment with the maximum voltage change rate, that is, the peak moment. Integrate the charging capacity between the earliest and the latest moments among all single batteries in the battery system up to this moment. The smaller the integration result, the better the consistency of the target single batteries.
[0048] Example: Assume that cell A in the system reaches its peak value earliest, corresponding to time Ta, and cell B reaches its peak value latest, corresponding to time Tb. After intercepting the data from Ta to Tb from the original data for processing, the charging capacity of the battery during this time interval is integrated, and the integrated result is used to evaluate the consistency of the battery system.
[0049] System Example: A lithium iron phosphate battery consistency calculation system includes a processor. The lithium iron phosphate battery consistency calculation system of this embodiment executes the lithium iron phosphate battery consistency calculation method as described in the method embodiment. The lithium iron phosphate battery consistency calculation method has been described clearly enough in the method embodiment and will not be repeated here.
Claims
1. A method for calculating consistency of lithium iron phosphate batteries, characterized in that: include: Charging, between two plateau periods of the target single battery's voltage changing with time, calculating the time interval between the moment when the single battery that reaches the largest voltage change rate earliest reaches the maximum corresponding voltage change rate and the moment when the single battery that reaches the largest voltage change rate latest reaches the maximum corresponding voltage change rate; integrating the charging capacity of the target single battery within the time interval; The smaller the integral result is, the better the consistency between the target single cells is.
2. The lithium iron phosphate battery consistency calculation method according to claim 1, characterized in that: The charging is constant current or constant power.
3. The lithium iron phosphate battery consistency calculation method according to claim 1, characterized in that: The moment when the voltage changes at the maximum rate is determined by the following steps: calculate the slope of the curve corresponding to the voltage changing with time, find the peak value of the slope that is not empty at several moments before and after as the target point, and take the moment when the slope is greater than the target point within the left and right set range as the moment when the voltage changes at the maximum rate.
4. The lithium iron phosphate battery consistency calculation method according to claim 3, characterized in that: If multiple target points satisfy the condition that the slope is greater than the slope within the left and right set ranges, the average of the maximum and minimum moments of these target points is taken as the moment when the voltage change rate is the largest.
5. A method for calculating consistency of lithium iron phosphate batteries, characterized in that: include: During charging, between the two plateau periods when the voltage of a single cell changes with time, calculate the time interval between the moment when the cell that first reaches the maximum voltage change rate reaches the maximum corresponding voltage change rate and the moment when the cell that last reaches the maximum voltage change rate reaches the maximum corresponding voltage change rate among all the single cells in the battery system; integrate the charging capacity of the battery system within the time interval; the smaller the integral result, the better the consistency of the battery system.
6. The lithium iron phosphate battery consistency calculation method according to claim 5, characterized in that: The charging is constant current or constant power.
7. The lithium iron phosphate battery consistency calculation method according to claim 5, characterized in that: The moment when the voltage changes at the maximum rate is determined by the following steps: calculate the slope of the curve corresponding to the voltage changing with time, find the peak value (extreme value) of the slope that is not empty at several moments before and after as the target point, and take the moment when the slope is greater than the target point within the left and right set range as the moment when the voltage changes at the maximum rate.
8. The lithium iron phosphate battery consistency calculation method according to claim 7, characterized in that: If multiple target points satisfy the condition that the slope is greater than the slope within the left and right set ranges, the average of the maximum and minimum moments of these target points is taken as the moment when the voltage change rate is the largest.
9. A lithium iron phosphate battery consistency calculation system, comprising a processor, characterized in that: The processor executes a computer program to implement the lithium iron phosphate battery consistency calculation method as described in any one of claims 1-8.