Turbo-shaft engine's controller design based on the compensation from disturbance was mainly studied in this article.A torque prediction model design scheme was proposed based on recursive reduced least squares support vector regression(RR-LSSVR)algorithm and model prediction procedure.Firstly

reduced least squares support vector regression was utilized to train the dynamic estimation nonparametric model of torque by using the handling input and flight state variables as training data.Then

a prediction procedure was arranged to obtain the torque information in advance.After that

the turbo-shaft engine's compensator was implemented using the former predicted torque information.In order to take full advantage of the disturbance signal and its dynamic information

the proportional-differential form was adopted in the compensator design.Finally

simulation of integrated helicopter/turbo-shaft engine system shows that the control scheme has better dynamic disturbance rejection ability than traditional system

it can decrease the droop of power turbine's speed more than 10%.