In　the　photovoltaic　power　generation　plant,　health　condition　of　each　unit　is　regarded　as　a　crucial　index.　Whether　the　life　prediction　is　accurate　or　not　determines　the　economic　benefit　of　the　plant.　Among　the　current　research　and　applications,　either　the　designed　system　is　too　complex　to　be　suitable　for　most　of　the　cases,　or　lack　of　complete　and　scientific　algorithm　to　support　the　analysis.　In　order　to　deal　with　the　problems　encountered,　in　this　paper,　a　life　prediction　method　for　photovoltaic　power　plant　based　on　GM(l,　l)　model　is　proposed.　On　the　basis　of　PR　calculation,　the　definition　of　Health　with　only　four　indicators　for　each　unit　is　introduced,　simplifying　system　calculation　and　improving　the　efficiency.　According　to　the　＇　＇Bathtub　Curve＇　of　the　equipment,　relate　aging　rate　to　Health.　After　calculating　and　transforming　the　collected　data　into　sequence,　the　preparatory　work　for　this　method　is　completed.　Further,　take　the　aging　rate　sequence　as　the　input　and　obtain　the　life　prediction　curves.　This　paper　designs　optimization　for　GM(l,　l)　grey　forecast　model,　which　makes　the　model　self-correct　the　curves　dynamically.　This　method　is　implemented　with　the　data　from　a　photovoltaic　power　plant.　The　results　show　that　this　method　is　well　worth　being　adopted　in　reality.　©　2020　IEEE.