Research On SOC Estimation And Equalization Control Strategy Of Lithium Battery Pack

With the increasingly serious fossil energy shortage and environmental pollution,the electric vehicles(EV)has been supported greatly by the state in recent years.As the core component in EV,the cell is the focus of research in academia and industry.Today,lithium cell have been widely used in EV.However,its voltage and capacity are limited,and a battery pack usually contains a large number of cells connected in series and parallel to meet the actual needs.Due to the difference of manufacturing technology and operating environment,cells always have difference that cannot be eliminated.A large number of cells and inconsistency bring a great challenge to EV battery management system(BMS),it mainly reflected in the following three aspects:(1)The accuracy of battery model is the basis of state-of-charge(SOC)estimation,battery equalization and other battery management work.However,it is difficult to describe comprehensively and accurately the characteristics of each cell because the inconsistency factors among cells are complex.It may lead to model mismatch and influence the estimation of important parameters further.(2)The state-of-charge(SOC)is one of the key parameters in the BMS.However,the resources of the on-board BMS are limited,and it is difficult to estimate the real-time SOC accurately for a large number of cells in the battery pack.(3)The inconsistency of available capacity among cells may reduce the overall available capacity of the battery pack.It is still a problem that using the effective balancing method to improve the inconsistency of available capacity and the energy utilization efficiency.Based on the battery pack,this paper focuses on the shortcomings of the existing research and the above practical problems.Then,the battery pack model,SOC estimation,and balancing control strategy were studied in this thesis.It provides a theoretical reference for the battery management system to realize accurate and rapid SOC estimation and efficient battery equalization.The main research contents are as follows:(1)To improve model accuracy and prevent model mismatch,we established a battery pack model that can reflect the coulomb efficiency difference among cells.Firstly,based on the one-order RC model,a dynamic relationship between the battery current and the average voltage of the battery pack is established.Then,the mean-model is built to describe the overall characteristics of the battery pack.Secondly,we analyzed the major inconsistency factors among cells.To improve further the accuracy of the battery pack model,the coulomb efficiency difference,SOC difference,and internal resistance difference were considered in the difference-model based on the Rint model.Finally,the validity and accuracy of the battery pack model are verified by experiments.(2)To meet the demand of battery management and the limitation of actual resources,a low-complex battery pack SOC estimation method is proposed.It is able to achieve an accurate and rapid SOC estimation of each cell at a low computational cost.Firstly,based on the mean-difference model,high sensitivity parameters were determined to reduce the calculation burden of online identification.Then,the state and parameters of the system were combined into an augmented state vector,and a new state-space equation was established.Secondly,in order to obtain accurate initial SOC values at a lower computational cost,dual time-scale AEKF was used to estimate the system states of the mean-model and the difference-model in turn.Thirdly,the coulomb efficiency short-term estimation model was established,and the parameters were identified online by combining the observed values and the forgetting factor recursive least square.This model and realtime observation values are used to update coulomb efficiency.Then,it is combined with the initial SOC value and current value to iterate the SOC of each cell by the real-time update method of the battery pack SOC rapidly.Finally,the accuracy and complexity of the estimation method are verified by experiments(3)In order to improve the overall energy utilization efficiency of the battery pack under the adjacent cell-to-cell balancing topology,a balancing control method of the battery pack with low energy consumption was proposed.Firstly,we analyze the adjacent cell-tocell balancing topology.Then,the energy transfer model of equalizer is established from the perspective of topological structure.Secondly,the balancing control is divided into two parts.One is balancing path control,and the other is balancing current intensity control.Then,the SOC is used as the balancing variable,and a global balancing path control strategy is designed to optimize the energy transfer process and reduce energy consumption.The battery pack balancing system model is constructed through the balancing strategy and the equalizer model.Thirdly,the model predictive control is used to control the balancing current intensity.The system prediction model is established through the recursive balancing system model.Then,the objective function of battery balancing control problem is designed to adjust the relationship between balancing speed and cell temperature rise by the weighting factor.According to the actual engineering environment,we design the constraints of SOC,trunk current,and balancing current.The way of moving optimization is used to solve.Finally,the validity of the proposed method for reducing energy loss is verified by experiment.