Evaluation On The Dynamic Consistency Of Li-ion Power Battery

Li-ion power battery has been the solid foundation for technologyinnovation within power battery field with its unique discharge specific capacityand long cycle life. Cell batteries should be connected in series to be applied forlarge appliance, though the rapid development in crafts and management.However, the general performance of the formed battery pack may suffer aapparent decay after a long term application, due to the deterioration in theconsistency of the battery based on large quantities of research. A betterevaluation on the consistency of battery is the very basis to keep betterperformance of the battery, to extend the cycle life, as well as to give advantageto battery management system (BMS) with maintenance.Various experiments are conducted to get to essence of consistency ofbattery, as to analyze the performance index of power battery for the typical onesfor the current state of battery in this paper. State of charge, as well as SOC andclosed circuit voltage (CCV) are selected with the research results for theperformance index of consistency.The equivalent circuit model is established and improved with weightvictor A(m) is involved to embrace the discrepancy in the very battery of theformed battery pack in the paper. Square-Cubature-Kalman-Filter, combined withStrong-Tracking Filter (SCKF-STF) algorithm is involved for SOC predictionwith corresponding simulation and error analysis. To obtain SOC predictionresult with higher precision, multiple filter algorithm is designed ahead of theSCKF-STF algorithm to tackle with the involved noise with input data. Thecomparison simulation of SOC prediction is conducted with optimized SCKF-STF algorithm and improved model. The corresponding simulation result anderror analysis is conducted with the single SCKF-STF for the adtantage of theoptimized algorithm. The probability density function of F-distribution with principle of statics isinvolved to obtain the evaluation on consistency of battery with SOC and CCV asperformance index. The corresponding description for probability densityfunction is deducted with actual index of experimented battery, as well as theprobability density curve. A pre-designed trust zone can be settled on the curvewith designed warning value. The zone of which is applied to make comparisonwith the result from deducted function with experiment data to evaluate thegeneral consistency of the experimented battery.