Velocity variation particle swarm optimization based predictive control for aero-engine

For the speed control of aero-engine, a predictive control algorithm based on least square support vector machine (LS-SVM) model and velocity variation particle swarm optimization (V-PSO) is presented. LS-SVM predictive model is constructed and compared with RBF neural network predictive model, the receding horizon optimization is realized by V-PSO, and then the control law is obtained. The simulation of the speed control system of aero-engine indicates that the steady state control of aero-engine can achieve good performance under the presented algorithm. When deviated from design point, the parameter of the LS-SVM predictive model will be modified on-line with respect to its own strong learning and adaptive abilities, which guarantee the closed-loop control system maintains good performance and strong robustness.