A　dual　adaptive　fading　extended　Kalman　filter(DAFEKF)　algorithm　is　proposed　for　the　problem　of　low　accuracy　and　convergent　speed　of　state-of-charge　(SOC)　estimation.　The　algorithm　designs　an　observer　of　state-of-charge　for　the　power　battery,　and　the　measured　current　and　voltage　are　taken　as　input　and　observation　values　of　the　observer,　respectively.　Then　the　state　of　charge　of　a　battery　is　estimated　by　the　DAFEKF.　The　DAFEKF　bases　on　the　Kalman　algorithm,　and　adds　the　time-varying　fading　factor　to　reduce　the　influence　of　past　data　on　current　filtering　values　and　to　adaptively　adjust　the　covariances　of　the　process　noise　and　measurement　noise.　SOC　results　of　a　lithium　battery　obtained　using　the　proposed　DAFEKF　are　compared　with　those　obtained　using　the　extended　Kalman　filter　(EKF)　and　the　adaptive　extended　Kalman　filter　(AEKF),　and　the　comparison　shows　that　the　DAFEKF　method　provides　better　accuracy,　robustness　and　convergence,　and　the　SOC　error　of　the　proposed　method　is　less　than　2%.　©　2018,　Editorial　Office　of　Journal　of　Xi＇an　Jiaotong　University.　All　right　reserved.