Premium
State‐of‐charge estimation of power lithium‐ion batteries based on an embedded micro control unit using a square root cubature Kalman filter at various ambient temperatures
Author(s) -
Cui Xiangyu,
He Zhicheng,
Li Eric,
Cheng Aiguo,
Luo Maji,
Guo Yazhou
Publication year - 2019
Publication title -
international journal of energy research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.808
H-Index - 95
eISSN - 1099-114X
pISSN - 0363-907X
DOI - 10.1002/er.4503
Subject(s) - state of charge , kalman filter , battery (electricity) , lookup table , mean squared error , control theory (sociology) , interference (communication) , voltage , square root , extended kalman filter , power (physics) , computer science , engineering , electrical engineering , real time computing , mathematics , control (management) , physics , statistics , geometry , quantum mechanics , artificial intelligence , channel (broadcasting) , programming language
Summary The development of a novel method to estimate the state of charge (SOC) with low read‐only memory (ROM) occupancy, high stability, and high anti‐interference capability is very important for the battery management system (BMS) in actual electric vehicles. This paper proposes the square root cubature Kalman filter (SRCKF) with a temperature correction rule, based on the BMS of a common on‐board embedded micro control unit (MCU), to achieve smooth estimation of SOC. The temperature correction rule is able to reduce the testing effort and ROM space used for data table storage (189.3 kilobytes is much smaller than the storage of the MPC5604B, with 1000 kilobytes), while the SRCKF is adopted to achieve highly robust real‐time SOC estimation with high resistance to interference and moderate computing cost (68.3% of the load rate of the MPC5604B). The results of multiple experiments show that the proposed method with less computational complexity converges rapidly (in approximately 2.5 s) and estimates the SOC of the battery accurately under dynamic temperature condition. Moreover, the SRCKF algorithm is not sensitive to the high measuring interference and highly nonlinear working conditions (even with 1% current and voltage measurement disturbances, the root mean square error of the proposed method can be as high as 0.679%).