A fast sorting method for power batteries based on parallel shunt characteristics

Abstract

This invention presents a rapid sorting method for power batteries based on parallel current shunting characteristics. It includes feature recognition based on the single-cycle discharge curve of the battery cells and dynamic modeling based on an optimized and improved CM battery model to determine whether the battery is within its normal operating range, thus achieving rapid sorting. The proposed rapid sorting method for power batteries has achieved good rapid evaluation results for power batteries under different operating conditions. This method can quickly characterize the module capacity utilization efficiency of batteries with low computational cost, providing a low-computation, highly flexible rapid battery sorting method, and offering technical and mechanistic support for the parallel assembly of battery cells and the cascade utilization of retired batteries.

Description

Technical Field

[0001] The present invention belongs to the field of electrical engineering, and in particular is a method for quickly sorting power batteries based on parallel shunt characteristics. Background Art

[0002] Lithium-ion power batteries have become the mainstream energy supply solution for new energy vehicles due to their excellent power density, long cycle life and other performance advantages. For the cell sorting and evaluation of retired batteries, the current traditional sorting mainly starts from the perspective of parameter differences to reduce the inconsistency between cells. However, this method has problems such as high time and computing costs and limited applicable scenarios. Ultimately achieving an improvement in module capacity utilization efficiency is a key indicator of the effectiveness of battery sorting methods. There is little research on battery dynamic modeling and sorting mechanisms based on efficiency improvement. Summary of the invention

[0003] The purpose of the present invention is to provide a fast battery sorting method with low computational cost and high flexibility, which provides technical and mechanism support for the parallel grouping of battery cells and the cascade utilization of retired batteries, including feature recognition based on the single-cycle discharge curve of the battery cell and dynamic modeling based on the optimized and improved CM battery model. The specific steps are as follows:

[0004] The present invention proposes a rapid sorting method for power batteries based on parallel shunt characteristics. The sorting method includes CM model establishment and a sorting process. The CM model establishment includes feature recognition based on a single-cycle discharge curve of a battery cell and dynamic modeling based on an optimized and improved CM battery model. The sorting process achieves rapid sorting by judging whether the battery is within a normal operating range. The specific steps are as follows:

[0005] (1) CM model establishment

[0006] (1.1) Feature recognition based on single-cycle discharge curve of battery cells

[0007] First, obtain the single cycle discharge curve of the battery cell during discharge. During the discharge process, the terminal voltage V of the battery cell is given by the following equation:

[0008]

[0009] Where: V is the battery terminal voltage, E0 is the no-load voltage, K is the polarization constant (V / Ah), Q is the rated capacity of the battery, R is the fixed internal resistance of the battery, I is the branch current, it is the battery discharge capacity, i * is the filter current, is the polarization voltage, is the polarization resistance, Aexp(-B·it) is the voltage drop at the initial stage of battery discharge;

[0010] Secondly, the single-cycle discharge curve is smoothed by filtering and noise reduction methods, the standard deviation of the data is calculated, and the range of the relatively stable part of the discharge voltage curve is obtained:

[0011]

[0012] Where: T is the sampling time of the battery discharge curve, U it is the terminal voltage of the ith cell, is the average voltage during the discharge of the battery cell; f rms is the data standard deviation;

[0013] (1.2) Dynamic modeling based on optimized and improved CM battery model

[0014] Take the data points (Q exp ,V exp )、(Q nom ,V nom ), the single cycle discharge curve is derivated to obtain the slope of the discharge curve, the starting point of the original discharge curve is obtained, and the third characteristic point (0, V full );

[0015] By identifying the data points at both ends of the single-cycle discharge curve and the third characteristic point, the dynamic modeling of the battery cell is completed based on these three characteristic value points:

[0016]

[0017] Where: V is the battery terminal voltage, V exp is the exponential region voltage, V nom is the cut-off voltage, E0 is the no-load voltage, K is the polarization constant (V / Ah), Q is the rated capacity of the battery, Q exp is the discharge capacity value in the exponential region, Q nom is the rated capacity of the battery, R is the fixed internal resistance of the battery, I is the branch current; I is the branch current, V full is the initial voltage, A is the voltage amplitude in the exponential region; B is the inverse of the time constant in the exponential region;

[0018] (2) In the sorting process, the CM model is used to calculate the ideal operating point between two battery cells to determine whether the battery is within the normal working range and achieve rapid sorting;

[0019] Select multiple battery cells with different parameters and connect them in parallel to form groups, and make different combinations of the battery cells to be sorted to determine the grouping type;

[0020] The CM model of formula (3) in step (1) is used to calculate the ideal operating point between the two battery cells. The load current corresponding to the distribution of the starting point current according to the capacity ratio is defined as the ideal operating point:

[0021]

[0022] Where: I i is the branch current of the ith cell, I L is the total branch current, R i is the DC resistance of the ith cell, C i is the rated capacity of the ith battery cell, U 0i is the initial voltage of the ith cell;

[0023] According to the calculated I Ideal , determine whether I is satisfied Ideal At 0.5I total and 1.5I total If the conditions are met, the grouping condition is met.

[0024] In the present invention, based on the dynamic discharge model optimized and improved by the CM battery model, only the discharge curve of a single cycle of the battery cell is used, the CM dynamic modeling is completed by identifying the characteristic value, the ideal working points between different battery cells are calculated, and rapid sorting is achieved.

[0025] In the present invention, the module capacity utilization rate after the cells are grouped is used to verify the rationality of the ideal working point of the rapid sorting method under multiple working conditions. The module capacity utilization efficiency is defined as:

[0026]

[0027] in: Refers to the SOC state of the ith cell when any cell in the module is discharged to the cut-off voltage, C i Refers to the rated capacity of the ith battery cell, C total Refers to the sum of the rated capacities of the modules;

[0028] When I1(0) / I2(0)=C1 / C2, SOC1(0+)=SOC2(0+), and Cell 1 and Cell 2 maintain the same SOC state. Maintain a constant state, so that the overall module capacity is fully utilized, where U OCVi is the open circuit voltage of the ith cell, SOC i The SOC state of the i-th battery cell.

[0029] The present invention is based on an optimized and improved dynamic discharge model of a CM battery model, only utilizes the discharge curve of a single cycle of a battery cell, completes CM dynamic modeling by identifying characteristic values, calculates the ideal operating points between different battery cells, and realizes rapid sorting.

[0030] The beneficial effects of the present invention are as follows: the present method provides a rapid battery sorting method with low computational cost and high flexibility, which can quickly characterize the module capacity utilization efficiency of the battery, and provide technical and mechanism support for the parallel grouping of battery cells and the cascade utilization of retired batteries. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 Schematic diagram of (a) battery CM model establishment and (b) battery sorting process.

[0032] Figure 2 Comparison of capacity utilization efficiency under four groups of different operating conditions; where: (a1) and (a2) are the capacity utilization rates under different operating conditions and working point ranges of #1, (b1) and (b2) are the capacity utilization rates under different operating conditions and working point ranges of #2, (c1) and (c2) are the capacity utilization rates under different operating conditions and working point ranges of #3, (d1) and (d2) are the capacity utilization rates under different operating conditions and working point ranges of #4. DETAILED DESCRIPTION

[0033] The present invention is further illustrated by the following examples.

[0034] Embodiment 1: According to Figure 1 As shown in the figure, a rapid sorting method for power batteries based on parallel shunt characteristics is proposed. This method can efficiently determine whether the battery is within the normal working range and achieve rapid sorting. First, the original discharge curve of a single cycle of the battery cell is filtered and smoothed, and the data points at both ends of the plateau period of the discharge voltage curve (Q exp ,V exp )、(Q nom ,V nom ), obtain the data within the standard deviation range in this interval, perform derivative processing on the discharge curve, obtain the slope of the discharge curve, obtain the starting point of the original discharge curve, and obtain the third characteristic point (0,V full ). By identifying the three characteristic values ​​of the discharge curve, the dynamic modeling of the battery cell is completed according to formula (3). According to formula (4), the ideal working points of different battery cells are calculated. By comparing whether the ideal working points fall within the normal working range of the module, it is evaluated whether the battery cells can be used for grouping, that is, to achieve the purpose of sorting.

[0035] First, select five cells with different parameters. The specific parameters are shown in the following table:

[0036]

[0037]

[0038] The battery cell was fully charged by constant current-constant voltage method, and after standing for 1 hour, it was discharged to the cut-off voltage at 0.5C, and finally the single-cycle discharge curve of the battery was obtained; then the five battery cells were connected in parallel to form four groups, which were numbered #1 to #4 respectively; a CM model was established for the discharge curve of each battery cell, and the working state corresponding to the highest efficiency of #1 to #4 was calculated as the ideal working point, and the module capacity utilization efficiency under different working conditions was calculated according to formula (5), as follows: Figure 2 As shown;

[0039] Depend on Figure 2 (a, c, e, g) It can be observed that the capacity utilization rate generally presents a normal distribution. The farther the working state point deviates from the highest efficiency, the lower the capacity utilization efficiency. When the ideal working point is in the range of 0.5C to 1.5C, the capacity utilization efficiency is close to 100%, which overall illustrates the effectiveness of using the ideal working point to evaluate the battery strategy.

[0040] Depend on Figure 2 (b, d, f, h) It can be observed that when the experiment is carried out with the ideal working point as the midpoint and 0.05C as the interval, it can be found that in the cases of (b2) and (c2), the ideal working point is not the point with the highest module capacity utilization efficiency. The reason can be explained by the branch current distribution under various working conditions.

[0041] The actual steady-state operation will have a slight deviation. Taking #2 as an example, in steady state, I1(∞) / C1>I3(∞) / C3, so if I3(0) / C3>I1(0) / C1 can be made up for in advance the overuse of Cell 1 capacity in steady state, the overall capacity utilization efficiency will be higher than I3(0) / C3=I1(0) / C1.

[0042] The sorting criterion proposed in the present invention is to determine whether the ideal working point falls within the normal working range (0.5C, 1.5C). If the ideal working point falls outside (0.5C, 1.5C), it means that the module has been operating at a low efficiency within the normal working range and is not suitable for grouping, thereby achieving the purpose of sorting. In this embodiment, the ideal working point of #1 is I ideal1 =0.21C total , which falls outside the normal operating range and is not suitable for grouping.

Claims

1. A rapid sorting method for power batteries based on parallel shunt characteristics, the sorting method comprising CM model establishment and a sorting process, the CM model establishment comprising feature recognition based on a single cycle discharge curve of a battery cell and dynamic modeling based on an optimized and improved CM battery model, the sorting process achieves rapid sorting by judging whether the battery is within a normal operating range; characterized in that The specific steps are as follows: (1) CM model establishment (1.1) Feature recognition based on single-cycle discharge curve of battery cells First, obtain the single cycle discharge curve of the battery cell during the discharge process. During the discharge process, the terminal voltage V of the battery cell is given by the following equation: Where: V is the battery terminal voltage, E0 is the no-load voltage, K is the polarization constant (V / Ah), Q is the rated capacity of the battery, R is the fixed internal resistance of the battery, I is the branch current, it is the battery discharge capacity, i * is the filter current, is the polarization voltage, is the polarization resistance, Aexp(-B·it) is the voltage drop at the initial stage of battery discharge; Secondly, the single-cycle discharge curve is smoothed by filtering and noise reduction methods, the standard deviation of the data is calculated, and the range of the relatively stable part of the discharge voltage curve is obtained: Where: T is the sampling time of the battery discharge curve, U it is the terminal voltage of the ith cell, is the average voltage during the discharge of the battery cell; f rms is the data standard deviation; (1.2) Dynamic modeling based on optimized and improved CM battery model Take the data points (Q exp ,V exp )、(Q nom ,V nom ), the single cycle discharge curve is derivated to obtain the slope of the discharge curve, the starting point of the original discharge curve is obtained, and the third characteristic point (0, V full ); By identifying the data points at both ends of the single-cycle discharge curve and the third characteristic point, the dynamic modeling of the battery cell is completed based on these three characteristic value points: Where: V is the battery terminal voltage, V exp is the exponential region voltage, V nom is the cut-off voltage, E0 is the no-load voltage, K is the polarization constant (V / Ah), Q is the rated capacity of the battery, Q exp is the discharge capacity value in the exponential region, Q nom is the rated capacity of the battery, R is the fixed internal resistance of the battery, I is the branch current; I is the branch current, V full is the initial voltage, A is the voltage amplitude in the exponential region; B is the inverse of the time constant in the exponential region; (2) In the sorting process, the CM model is used to calculate the ideal operating point between two battery cells to determine whether the battery is within the normal working range and achieve rapid sorting; Select multiple battery cells with different parameters and connect them in parallel to form groups, and make different combinations of the battery cells to be sorted to determine the grouping type; The CM model of formula (3) in step (1) is used to calculate the ideal operating point between the two battery cells. The load current corresponding to the distribution of the starting point current according to the capacity ratio is defined as the ideal operating point: Where: I i is the branch current of the ith cell, I L is the total branch current, R i is the DC resistance of the ith cell, C i is the rated capacity of the ith battery cell, U 0i is the initial voltage of the ith cell; According to the calculated I Ideal , determine whether I is satisfied Ideal At 0.5I total and 1.5I total If the conditions are met, the grouping condition is met.

2. The method for rapid separation of power batteries based on parallel shunt characteristics according to claim 1 is characterized in that: The dynamic discharge model optimized and improved based on the CM battery model only uses the discharge curve of a single cycle of the battery cell. By identifying the characteristic values, the CM dynamic modeling is completed, the ideal working points between different battery cells are calculated, and rapid sorting is achieved.

3. The method for rapid separation of power batteries based on parallel shunt characteristics according to claim 1 is characterized in that: The module capacity utilization rate after the cells are grouped is used to verify the rationality of the ideal working point of the rapid sorting method under multiple working conditions. The module capacity utilization efficiency is defined as: Among them: SOC remaini Refers to the SOC state of the ith cell when any cell in the module is discharged to the cut-off voltage, C i Refers to the rated capacity of the ith battery cell, C total Refers to the sum of the rated capacities of the modules; When I1(0) / I2(0)=C1 / C2, SOC1(0+)=SOC2(0+), and Cell 1 and Cell 2 maintain the same SOC state. Maintain a constant state, so that the overall module capacity is fully utilized, where U OCVi is the open circuit voltage of the ith cell, SOC i The SOC state of the i-th battery cell.

Images