Study On Lithium-ion Battery SOC Estimation And Battery Swapping With Energy Block

Lithium ion batteries are widely used in new energy vehicles due to lowering cost combined with excellent energy density as well as cycle durability. The need for real-time state of charge(SOC) estimation and study on battery pack mode arises from a demand to predict the residual mileage and optimize the energy output performance of the full pack. This paper carried out the following work on the two research hotspots.Based on the Lithium-ion battery pack, to estimate the battery SOC online, we first design the test scheme of the battery, an equivalent circuit model with one RC branch is built and the parameters are identified. The model accuracy is validated by three different dynamic loading profiles. Then four algorithms derived from Kalman Filter are compared. The comparison of the experimental results shows that the adaptive unscented Kalman filter(AUKF) has the best accuracy. Not only does it solve the problem that the accuracy overwhelmingly depends on the information of the noise, it also has the potential against typical uncertainties of electrical vehicle. But it exhibits huge error when the internal parameters of the battery are not certain.For the above problem, the accuracy of AUKF algorithm with uncertain capacity and with uncertain resistance and capacitance are analyzed, it turns out that the former uncertainty dose great impact on SOC estimation, and the later uncertainty dose on voltage. Therefore Parameter Varying Approach(PVA) is used to update resistance and capacitance online, the results suggests that the voltage estimation accuracy is effectively improved by the difference between the actual voltage and the estimated voltage. Afterwards, we develop an offline capacity estimation method based on OCV and SOC variation; Experimental validations demonstrate that with proper OCV range, the capacity of battery can be estimated accurately.Finally, based on the SOC of a single energy block, to optimize the energy output performance of the full battery pack, we discuss battery swapping with energy block. Single soft package battery is connected in series to form a energy block, a plurality of energy blocks are then connected in parallel to form a battery, users can put as many energy blocks in parallel as they need. 20 series 8 parallel battery pack model is built in Simulink to research various situations. Simulation results prove that battery modules with different SOC can be connected in parallel without adjusting difference in advance. They can be automatically balanced and the equilibrium current will not exceed the allowable range. Moreover, when a battery cell is short-circuited, if the total discharge current of the battery pack is large enough, the battery module in short circuit can still provide energy, avoiding the capacity reduction. The battery swapping with energy block has already had a batch application in a car factory.