On-line State Of Charge Estimation Of Power Lithium-ion Batteries

ABSTRACT:Under the background that energy shortage and environmental pollution problemsof the world have been increasingly serious, the electric vehicles become the hot topic of recent research for incomparable advantages in energy saving and environmental protection. The state evaluation problems of power lithium-ion batteries in electric vehicles, especially SOC (state of charge) estimation, always are the hotspots and difficulties of research in the field of the battery. At the same time, SOC estimation plays a key role in the energy management system. Besides, in the application of electric vehicles, power lithium-ion batteries are connected in groups, so the subject of this paper is on-line state of charge (SOC) estimation for lithium-ion battery packs. Details of the research and achievement are as follows:Firstly, the basic working principle and structure of the lithium-ion battery were analyzed, then temperature and aging characteristics for parameters, such as maximum discharging capacity, internal resistances and SOC-OCV curves, were also analyzed here, which builds the basis of state estimation model; On the basis of cell circuit models and PI state Observer principle, the SOC estimation model was established. Then state estimation error modeling and sensitivity analysis are focused on in the paper, while corresponding key influencing factors are obtained.Secondly, a general SOC-OCV model which is available for life span and whose coefficients are identifiable was proposed here, then model robustness characteristics were discussed and this new model was with good robustness; besides, recursive least squares methods for parameter on-line identifying and updating guaranteed SOC estimation accuracy.Thirdly, considering inconsistency of each monomer in the group, definition of the battery pack SOC was proposed based on the relationship between single cell state and SOC of battery pack. The simulation model for SOC estimation which combines state observer and ampere hour integration method was built and this method was verified through the test for LNMCO battery packs.Finally, a hardware-in-loop test platform for battery modules was established. On-line SOC estimation tests under different temperature and conditions for two different battery group modules were lauched and estimation precision of this method was also verified. In addition, involved experimental methods and conclusions in this paper are also available for other types of power lithium-ion batteries.