Magnetic　hyperthermia　destroys　malignant　cells　by　absorbing　the　heat　generated　by　magnetic　nanoparticles　(MNPs)　under　an　external　AC　magnetic　field.　One　of　the　most　influencing　factors　for　such　therapy　is　to　control　the　heat　generation　within　MNPs,　which　is　highly　relevant　to　applied　AC　magnetic　field　and　used　MNPs　during　treatment.　This　paper　optimizes　the　intensity　of　magnetic　field　acting　on　MNPs　by　considering　an　optimum　power　dissipation　of　MNPs,　which　is　obtained　by　using　Nelder-Mead　algorithm　for　a　proposed　objective　function　during　the　optimization　procedure.　The　aim　of　the　proposed　objective　function　is　to　minimize　the　difference　between　the　maximum　treatment　temperature　and　its　upper　safety　value　as　much　as　possible.　The　optimization　results　can　be　used　to　investigate　the　injury　degree　for　the　proposed　model　by　Arrhenius　thermal　damage　model,　and　also　to　optimize　the　performance　for　the　magnetic　field　strength　on　MNPs　by　the　implicit　function　due　to　Rosensweig＇s　theory　in　this　study.　Simulation　results　demonstrate　that　the　power　dissipation　of　MNPs　can　quickly　converge　to　an　optimum　value　after　less　iterations　during　the　optimization　process,　which　can　also　result　in　an　ideal　treatment　temperature　profile　and　can　further　obtain　the　injury　degree　for　different　malignant　cells,　when　a　temperature-dependent　blood　perfusion　rate　is　considered　for　the　simulation.　The　magnetic　field　or　the　MNPs　can　also　be　optimized　before　the　therapy　under　this　optimal　power　dissipation　for　MNPs　at　this　time.　In　addition,　this　study　also　reveals　the　evolution　of　treatment　temperature　due　to　two　blood　perfusion　rates,　namely　temperaturedependent　and　constant　models.