The Modeling And Charging Strategy Optimization For Liquid Metal Batteries

Energy storage is an effective solution to solve the problem of renewable energy integration and improve the utilization rate of renewable energy.As a new energy storage technology,liquid metal batteries(LMBs)have a broad application prospect in the large-scale energy storage field for power grid with the advantages of low cost,large capacity and long life.The safe and efficient operation of batteries is inseparable from advanced battery management system(BMS).It is urgent to study the key theories and technologies of BMS while the energy storage properties of LMBs are quite different from conventional lithium ion batteries.In order to describe the dynamic characteristics of LMBs accurately and improve the performance of batteries,a series of work is carried out on battery modeling and charging strategy optimization based on the basic battery characteristics and testing methods,so as to provide guidance for safe and efficient management of LMBs.The main research contents and results are summarized as follows:1.Open circuit voltage(OCV)characteristic,rate characteristic and internal resistance(IR)characteristic of LMBs are studied and the influence of state of charge(SoC)on the IR is analyzed from the aspects of reaction interface evolution and mass transfer process.The results show that the LMBs have wide OCV platform and remarkable rate characteristics.Meanwhile,the ohmic resistance is about 12 mΩ and the concentration polarization resistance in discharging process is greater than charging process.2.There common equivalent circuit models are compared and the dual-polarization model is selected considering model accuracy and computational complexity.Based on the data of hybrid pulse power characterization(HPPC)tests,model parameters are identified using least square method.Battery cycle tests and HPPC tests are carried out at different working temperatures.The influence of working temperature on OCV,charge-discharge performance and IR is analyzed.Dual-polarization model considering temperature effect is built and simulation results have proved the error is within 0.03 V.3.A dual-equivalent circuit fusion model is proposed in which Rint model and Thevenin model are used to characterize the dynamic characteristics of LMBs in long-time scale and short-time scale,respectively.Simulation results have proved the proposed model has high model accuracy under both constant-current and variable-current conditions.4.A multi-stage constant current(CC)charging method is proposed which utilizes adaptive current based on IR to reduce charging loss.CC pulse tests are carried out to obtain IR curve and ordinal sample cluster is used to segment the IR curve.Genetic algorithm is applied to search the optimal current sequence and simulation results have proved that the proposed charging stratege decreases the charging loss by 0.13%～0.16%.5.Constant current constant voltage(CCCV)cycle tests under differert charging rate,cut-off voltage and constant-voltage time are carried out and the effect of cut-off voltage and constant-voltage time on battery performance is analyzed.A multi-objective optimization model considering charging time,discharging capacity and energy efficiency is built.The CCCV charging method with optimized parameters increases discharging capacity and energy efficiency while reducing charging time compared with the traditional charging method.