The　return　voltage　measurement　can　effectively　detect　the　moisture　content　of　solid　insulation　in　transformer,　but　it　is　difficult　to　get　the　definite　diagnostic　results　from　the　return　voltage　polarization　spectrum　with　local　peaks.　The　theory　of　extended　Debye　model　is　introduced,　the　relationship　among　insulation　polarization,　aging　process　and　equivalent　circuit　parameter　is　established,　and　the　variation　law　of　characteristic　variables　and　polarization　spectrum　induced　by　the　variation　of　electric　parameters　are　simulated　and　analyzed.　Simulative　results　show　that,　the　dominant　time　constants　can　diagnose　the　insulation　conditions　of　transformers　with　different　geometric　capacitances,　which　is　an　effective　characteristic　variable.　The　shape　of　polarization　spectrum　is　also　an　effective　characteristic　variable　of　insulation　diagnosis.　The　shape　of　its　terminal　part　and　its　return　voltage　can　reflect　the　aging　condition　of　solid　insulation.　Test　results　validate　the　effectiveness　of　simulation　conclusions.