As　an　important　form　of　power-by-wire　(PBW)　technology,　electro-hydrostatic　actuator　(EHA)　is　widely　used　in　marine　ships　and　aircrafts　because　of　the　characteristics　of　high　performance,　high　reliability　and　high　power　density　in　the　future.　However,　the　current　EHA　applied　for　ship　bilge　pumps　has　the　problem　of　significantly　increase　in　response　time.　A　data-based　performance　degradation　model　of　EHA　is　proposed　in　this　paper,　which　considering　the　physical　model　and　actual　on-line　data.　Simulink　block　diagram　model　is　carried　out　to　simulate　the　EHA　under　actual　working　conditions.　EHA　performance　degradation　is　simulated　through　parameterized　modeling.　A　backpropagation　artificial　neural　network　(BP　ANN)　is　applied　to　simulate　the　performance　degradation　of　on-line　EHA,　which　can　be　furtherly　used　for　prognostics　and　health　management　(PHM)　of　EHA　under　actual　working　conditions.　The　results　of　simulation　and　experiments　show　that　the　proposed　method　can　well　simulate　the　response　time　with　performance　degradation　model　and　on-line　data,　which　is　significant　for　the　real-time　PHM　of　EHA　system.