The　optimal　design　of　the　mobile　crossbeam　falling　motion　control　system　is　an　important　way　to　improve　the　stability　of　the　composite　hydraulic　press.　At　present,　the　falling　control　strategy　is　mostly　designed　by　the　empirical　method,　which　has　the　disadvantages　of　low　efficiency　and　long　time.　In　this　paper,　a　nonlinear　hydraulic　control　system　model　and　three　types　of　falling　trajectory　mathematical　models　are　established　for　the　mobile　crossbeam　falling　motion　control　system.　Based　on　the　comparison　of　the　motion　characteristics　of　each　trajectory,　a　cubic　polynomial　falling　trajectory　which　can　effectively　suppress　vibration　is　designed.　This　paper　proposes　a　method　of　trajectory　planning　based　on　the　spline　interpolation　function　and　optimizes　the　trajectory　globally　with　an　improved　migration　genetic　algorithm;　the　optimized　system　ITAE　index　is　improved　by　49.56%　compared　with　the　uniform　speed　trajectory.　The　experimental　results　show　that　the　optimized　trajectory＇s　vibration　settling　time　is　reduced　by　62.91%　and　the　maximum　pressure　fluctuation　is　reduced　by　25.34%　compared　with　the　uniform　speed　trajectory.