The　paper　proposes　an　ultra-local　model　predictive　current　control　strategy　to　improve　the　stability　and　robustness　of　the　phase-shifted　full-bridge(PSFB)　converter　under　parameter　mismatch　causing　by　operating　conditions　changes.　Firstly,　analysis　is　conducted　to　show　the　effect　of　circuit　parameter　mismatch　on　the　anticipated　current　estimate　in　the　PSFB　converter　circuit.　Secondly,　an　ultra-local　model　is　built　for　the　PSFB　converter　by　analyzing　the　total　disturbance　of　the　circuit.　Ultimately,　the　unknown　nonlinear　total　disturbance　part　of　the　model　is　estimated　using　a　Kalman　filter,　which　is　then　integrate　in　model　predictive　control　algorithm　to　enhance　the　control　effect.　The　experiments　on　a　semi-physical　simulation　platform　demonstrate　that　the　proposed　strategy　effectively　suppresses　the　steady-state　error　caused　by　parameter　mismatch　when　compared　to　the　conventional　predictive　current　control　scheme.　It　increases　system　resilience　and　anti-disturbance　capabilities　in　the　event　of　parameter　disturbances,　as　well　as　its　capacity　to　adapt　efficiently　to　changes　in　working　conditions.