A　method　of　fuselage　shape　control　based　on　digital　twin　drive　was　proposed　in　order　to　solve　the　problems　of　low　precision,　low　efficiency　and　excessive　local　stress　of　manual　adjustment.　A　digital　twin　system　integrating　shape　control　strategy　optimization　algorithm　and　virtual　debugging　technology　was　constructed.　The　digital　twin　system　was　driven　by　real-time　data.　The　data　interaction　and　the　dynamic　mapping　between　physical　space　and　virtual　space　of　fuselage　shape　control　system　were　realized.　The　optimization　problem　of　shape　control　strategy　combining　genetic　algorithm　and　deep　learning　was　analyzed.　Then　the　availability　of　shape　control　strategy　was　verified　through　the　ANSYS　batch　multi-load　step　method.　The　shape　of　fuselage　was　adjusted　by　maintaining　the　stress　balance　of　fuselage　barrel　section.　The　experimental　results　show　that　the　digital　twin　system　and　shape　control　strategy　can　effectively　improve　the　control　accuracy　of　fuselage　barrel　section　shape　by　25.8%,　improve　the　control　efficiency　of　fuselage　barrel　section　shape　by　414.3%　and　reduce　the　local　maximum　stress　by　42.5%.　©　2022　Zhejiang　University.　All　rights　reserved.