Simulation of a single battery may be complicated; simulation of a battery pack of multiple batteries is more difficult and tedious.
Most studies have focused on a single battery, very little were done on multiple batteries, especially lithium-ion batteries.
Battery packs of multi-batteries supply high voltage when batteries are connected in series and high capacity when connected in parallel. Generally in battery packs of parallel connection, batteries of similar parameters (type, resistance, and capacity) are chosen. Gan and Takeuchi have shown that a battery of high discharge capability with a battery of large capacity in parallel has better discharge performance and better discharge efficiency than any of the single battery Numerical simulation for the discharge behaviors of batteries in series and/or parallel-connected battery pack Mao-Sung Wu , Chang-Yen Lin , Yung-Yun Wang , Chi-Chao Wan , C.R. Yang A calculation method has been proposed based on a semi-experimental equation Discharge data of different types of single lithium- ion battery has been used as input Discharge behavior of the entire battery pack composed of several different-type batteries in series and/or in parallel may be calculated Every single battery in a pack may be monitored.
Discharge curve of lithium ion batteries of 750 mAh, 1300 mAh, 1900 mAh and 2000 mAh at different discharge currents have been collected Voltage variation prediction in constant current discharge can be made A new discharge curve at another current can be predicted which is helpful for predicting the behavior similar batteries in series and parallel.
Similar batteries in series will have the same current of the entire system(Battery Pack) System voltage is the sum of each individual voltages In parallel connection the systems current is equally distributed to each battery for similar type of batteries. Practically parallel connection is more complex for batteries of different discharge capacities.
A iteration method is used. For two batteries for example A and B in parallel I B (t)=I system (t)-I A (t)
After I A (t) is preset I B (t) can be calculated to obtain their corresponding voltage. This iteration continues with a view to getting same voltage across both the batteries. Binary search in data structures is common for picking correct current value. Convergence condition is set at |VA-VB|<10 8 Calculation Method for a multi-battery system in parallel and/ or series The discharge behavior of the pack can be measured, regardless whether the batteries are similar or different Multi battery simulation is required any circuitry may be simplified into parallel- or series-connections of two batteries Three battery circuitry is discussed
For series connection V system (t)= V A (t)+ V B (t)+ V C (t) V A (t)=V A (I system (t), C A (t)) V B (t)=V B (I system (t), C B (t)) V C (t)=V C (I system (t), C C (t)) Current is given as input and voltage across each battery can be found.
In case of parallel connection I system (t)=I A (t)+ I B (t)+ I C (t) V system (t)=V A (t) V system (t)=V B (t) V system (t)=V C (t) V A (t)=V A (I system (t), C A (t)) V B (t)=V B (I system (t), C B (t)) V C (t)=V C (I system (t), C C (t))
Iteration is used to estimate the currents Here I A (t) and I B (t) should be preset to get I C (t) Time consuming method Simplification has been done I system (t)>= I A (t)+I B (t) as I C (t) cannot be to extreme(I system (t) or zero)
Parallel connection between A and B has been considered and discharge curves between I system (t) and zero have been extracted. A and B is considered as a new battery D. Now I system (t)=I C (t) + I D (t). Current for C and D are calculated Then current for A and B are calculated from I D (t). A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Difference among the cells and impact of balance control have been considered Series and parallel connection are studied Unsccented particle filter algorithm has been used Based on the model of the first overcharged cell and that of the first over-discharged cell, the estimation of the SOC of a battery pack is realized Experimentally verified that an accurate estimation value can be obtainedan accurate estimation value can be obtained A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen The in-pack cell with the lowest available capacity will determine the available capacity of the entire pack The charging of the pack will stop when the in- pack cell with the highest available capacity is full The variations in the individual cells are primarily the cell total capacity, internal resistance and the initial value of the SOC A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen A battery management system (BMS) with a balance control unit addresses the uneven cell problem In this study, the relationship between the pack parameters (SOC and total capacity) and the parameters of the in-pack cells with consideration of the different balance control strategies has been established A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen The SOC of B1 and B2 are determined by the UPF algorithm with their corresponding battery models, and then the pack SOC can be calculated based on their relationship. B1=First Over discharged cell B2=First over charged cell Definition of the pack SOC is first described The relationship between the parameters of the pack and those of the in-pack cells has been analyzed A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Common Battery model for B1 and B2 The UPF algorithm is applied to estimate the SOC of B1 and B2 Using the parameters of B1 and B2, the pack SOC is calculated SOC of a single cell A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen The relationship between battery units and the pack can be converted into the relationship between the in-pack cells and the pack
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Series Connection The SOC and total capacity of each cell are denoted by SOC i and C i Without Balance Control
a)The battery pack capacity is the sum of the minimum cell capacity that can be charged and the minimum remaining cell capacity that can be discharged
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen b) The total capacity of the pack
c) SOC of the Pack
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen c) The minimum remaining cell capacity in the fully charged pack as c min , the corresponding cell SOC as SOC cell and the total capacity of the cell as C cell The SOC of the pack is
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Passive Balance Control a) Pack total capacity
a) The pack SOC with passive control is
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Active Balance Control a) Charge is transferred between cells to achieve the balance b) Total capacity of the pack is the mean capacity of the cells
c) SOC of the pack is
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Parallel Connection Each branch can be regarded to be a one unit, with the SOC and total capacity of each branch denoted by SOC pack _i and C pack_i Total Capacity of the pack is
SOC of the pack is
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Cell with the minimum chargeable capacity is first to be over-charged(B2) Minimum remaining capacity is first to be over- discharged(B1) pack SOC can be simplified by the parameters of B1 and B2
Passive Balance: cell with the minimum total capacity is first to be over-discharged, so the pack SOC can be expressed as the SOC of B1
Active Balance: The pack SOC can be the SOC of any one of the cells SOC in the pack. To reduce the estimation error, the pack SOC is represented by the mean SOC of B1 and B2
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen SOC and total capacity of each branch can be calculated as mentioned below.
SOC and total capacity of each branch will lead to the calculation of SOC and capacity of total parallel pack. A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen Lithium Ion Battery Model 1. Single Battery Model is
2. To estimate the SOC of B1 and B2, their respective models are constructed 3. Linear state space equations are derived through discretization
A method for the estimation of the battery pack state of charge based on in-pack cells uniformity analysis Liang Zhong, Chenbin Zhang, Yao He, Zonghai Chen UPF algorithm has been used to estimate the single cells SOC and the pack SOC based on the non-linear state space equations of B1 and B2. Simulation and Experimental results I. To analyze the impact of the cell-to-cell variation on the estimation of the pack SOC II. To verify the applicability of the algorithm based on UPF III. The pack is composed of IPF1865140 type cells in series with a nominal capacity of 10 Ah. The cells are with a nominal voltage of 3.2 V
Battery Model Identification and validation has been done i.e. parameters of the battery model were determined experimentally and compared Analysis of the UPF algorithm and the estimation of the pack SOC a. the battery pack was first discharged at C1 rate for approximately 20 min and then operated for approximately 10 cycles b. The error of the estimation of a single battery SOC based on the UPF algorithm is 1.0779%, while the error based on the UKF algorithm is 2.803% Cell SOC Estimation Curve for two methods C. With the choice of the appropriate battery model, the SOC of the first over-discharged battery B1 and the first over- charged battery B2 could be obtained by using the UPF algorithm
D. For the case when the initial value of the SOC of each in- pack cell is different from each other, the SOC estimation of the pack was analyzed. Assume the SOC of B1 is 50% and B2 is 55%. SOC estimation error of B1 and B2 were 1.0789% and 1.3903% respectively Cell SOC Estimation with different initial value SOC Estimation of a battery Pack with inconsistency of the initial SOC value among in-pack cells SOC Estimation of a battery Pack with inconsistency of the initial SOC value among in-pack cells Errors for different types balance are calculated
SOC Estimation of cells with different internal resistance and same Initial SOC for cells The SOC estimation of the cells with different total capacity and their SOC comparison Assumption(B1-50% and B2-55%) The SOC estimation of the battery pack with inconsistency of the in-pack cell total capacity Conclusion The pack SOC estimation was transformed into the estimation of the parameters of the first over- charged cell and the first over-discharged cell. Applying the UPF algorithm to the single cellmodel, the SOC of the single cell was obtained, and thus the pack SOC was calculated Parallel connection was not studied