Lithium-ion　batteries　have　been　widely　used　in　various　electronic　and　electrical　equipment,　and　the　prediction　of　their　remaining　useful　life　is　essential　to　ensure　the　safe　and　reliable　operation　of　the　system.　To　solve　the　problem　of　predicting　the　remaining　useful　life　of　lithium-ion　batteries　under　the　condition　of　missing　capacity　values,　a　method　based　on　the　fusion　of　neural　networks　and　stochastic　degradation　models　is　proposed.　First,　the　artificial　bee　colony　algorithm　is　used　to　automatically　adjust　the　structural　parameters　of　the　echo　state　network,　and　a　capacity　prediction　framework　based　on　ABC-ESN　is　constructed　to　predict　the　capacity　value　of　the　lithium-ion　battery　at　the　next　moment　and　other　moments　in　the　future.　Then,　an　adaptive　two-stage　stochastic　degradation　model　is　established　to　derive　the　probability　density　function　of　the　remaining　useful　life　of　the　lithium-ion　battery　at　the　next　moment　and　other　moments　in　the　future.　Finally,　all　the　model　parameters　and　the　remaining　useful　life　distribution　of　the　lithium-ion　battery　at　the　next　moment　and　later　moments　are　updated　by　combining　the　predicted　capacity　value　and　unscented　particle　filter　algorithm　to　realize　the　prediction　of　the　remaining　useful　life　in　advance.　The　results　of　prediction　experiments　verify　the　effectiveness　and　accuracy　of　the　proposed　method.　　©　2022　IEEE.