Air-coolant　proton　exchange　membrane　fuel　cells　(PEMFCs)　have　the　characteristics　of　compact　structure　and　low　cost,　benefiting　for　lightweight　applications　such　as　aircrafts　with　limited　volume.　However,　the　air-coolant　PEMFC　has　strong　coupling　between　air　supply　and　cooling　resulting　in　difficultly　analyzing　the　relationship　between　optimal　power　and　stack　temperature　to　cause　low　power　efficiency.　To　effectively　increase　the　output　power　of　the　air-coolant　PEMFC　under　changeable　environmental　conditions,　this　article　introduces　a　real-time　power　optimization　strategy　based　on　the　active　temperature　control.　To　output　the　continuously　maximum　power,　an　improved　temperature　perturb　and　observe　(P&O)　is　designed　to　obtain　an　optimal　temperature　reference.　Considering　the　PEMFC　is　a　highly　nonlinear　system,　the　super-twisting　algorithm　(STA)-based　controller　is　developed　to　regulate　the　stack　temperature.　After　the　optimal　temperature　is　determined,　the　STA　controller　may　lead　to　the　target　temperature　be　well　tracked　with　strong　robustness　under　unknown　external　disturbances.　To　validate　the　proposed　strategy,　practical　experiments　were　carried　out　in　a　1-kW　air-coolant　PEMFC　under　different　operation　scenarios,　and　the　results　showed　that　the　power　was　increased　over　4%　when　the　air-coolant　PEMFC　operated　under　relatively　high　load.　Overall,　the　proposed　strategy　looks　promising　for　power　optimization　of　air-coolant　PEMFCs.　©　2020　Elsevier　Ltd