To　accurately　predict　the　state　of　health　(SOH)　of　lithium-ion　batteries　and　improve　the　safety　and　reliability　of　battery　management　systems,　a　new　SOH　estimation　method　based　on　fusion　health　features　(HFs)　and　adaptive　boosting　integrated　grey　wolf　optimizer　to　optimize　back　propagation　neural　network　(Adaboost-GWO-BP)　is　proposed.　First,　five　kinds　of　multi-type　HFs　were　extracted　from　the　battery　charging　process,　and　the　correlation　between　the　proposed　HFs　and　SOH　was　verified　by　Pearson　and　Spearman　correlation　coefficients.　Then,　the　indirect　health　feature　(IHF)　was　obtained　by　multidimensional　scaling　dimensionality　reduction　to　reduce　data　redundancy　and　improve　the　correlation　between　HFs　and　SOH.　The　GWO-BP　model　was　then　used　to　establish　the　nonlinear　mapping　relationship　between　IHF　and　SOH.　In　order　to　overcome　the　problem　of　low　accuracy　of　battery　SOH　estimation　in　a　single　model,　the　Adaboost　algorithm　in　ensemble　learning　is　introduced　to　enhance　the　accuracy　of　the　model　estimation.　Finally,　the　proposed　method　is　verified　by　NASA　dataset,　and　compared　with　other　models.　In　the　comparative　experiments,　mean　absolute　error　and　root　mean　square　error　of　the　proposed　method　for　SOH　estimation　is　less　than　0.81%　and　1.26%,　which　has　higher　accuracy　compared　to　other　models.　(c)　2024　The　Electrochemical　Society　(＂ECS＂).　Published　on　behalf　of　ECS　by　IOP　Publishing　Limited.　All　rights,　including　for　text　and　data　mining,　AI　training,　and　similar　technologies,　are　reserved.　[DOI:10.1149/1945-7111/ad940c]