Aiming　at　the　problem　of　uncertainty　in　parameter　solution　of　relaxation　response,　the　indirect　extraction　of　feature　quantities　and　the　difficultly　in　describing　physical　meaning　in　the　relaxation　response　parameters　extraction　for　the　time　domain　dielectric　response　method,　a　method　for　diagnosing　the　aging　state　of　transformer　insulating　paper　and　insulating　oil　based　on　direct　deterministic　parameter　identification　and　characteristic　parameters　extraction　was　proposed.　Firstly,　a　terminal　two-point　analytical　method　was　used　to　determine　the　relaxation　parameter　in　the　depolarization　current　spectrum.　The　relaxation　minimum　spectrum　and　maximum　spectral　spectrum　which　characterize　the　state　of　transformer　insulating　oil　and　insulating　paper　were　drawn　separately,　and　four　new　feature　quantities　were　directly　extracted:　relaxation　minimum　normal　angle　θτmin,　relaxation　minimum　normal　length　Pτmin,　relaxation　maximum　normal　angle　θτmax,　relaxation　The　maximum　normal　length　Pτmax.　It　has　been　confirmed　by　research　and　analysis　that　the　more　severe　the　aging　of　insulating　oil　is,　the　smaller　the　θτmin　and　shorter　the　Pτmin,　the　more　severe　the　aging　of　insulating　paper,　the　smaller　the　θτmax　and　shorter　the　Pτmax.　Test　verification　of　multiple　transformers:　according　to　the　new　feature　quantity,　the　aging　state　of　oil　and　paper　insulation　can　be　effectively　diagnosed;　the　new　method　of　relaxation　parameter　identification　and　new　feature　quantity　extraction　provide　a　new　idea　for　oil-paper　insulation　aging　diagnosis.　©　2020,　Harbin　University　of　Science　and　Technology　Publication.　All　right　reserved.