The support vector machine principal and the support vector machine return model are introduced first

and the self-adaptive PID support vector machines' controller is established.It makes use of support vector machines' strong nonlinear mapping ability

and network's structure is optimally auto-created.Finally

the aeroengine controller is designed at the design point.Application results in an aeroengine show that the controller possesses much better performance when the work of the aeroengine deviates from the design point.The presented control strategy introduces a new way for nonlinear control.