基于LS-SVM的非线性鲁棒自适应飞行控制器设计

Nonlinear robust adaptive flight controller design based on LS-SVM

为解决因参数不确定,受外界未知扰动等非线性影响的飞行控制问题,以反馈线性化理论为基础,将最小二乘支持向量机引入控制结构,设计了一种基于在线逆误差补偿的非线性鲁棒自适应飞行控制器。经分析推导得出最小二乘支持向量机的自适应权值调整率,并应用Lyapunov稳定性理论证明飞行控制系统闭环渐进稳定。针对某型飞行器,通过构建平均气动模型误差方程模拟实际飞行环境下的不确定非线性影响,开展了6自由度飞行仿真对比试验,仿真结果证明所设计的控制器具有良好的指令跟踪能力和较强的鲁棒性。

A nonlinear robust adaptive flight controller based on feedback linearization theory is proposed to solve the flight control problems which are induced by nonlinear problems, such as uncertain parameters or disturbances. And a least square support vector machine （LS-SVM） is introduced in control scheme to compensate the inverse error online. The updating rule of LS-SVM parameters is derived, and the closed-loop nonlinear system is guaranteed to be asymptotically stable by Lyapunov theory. Furthermore,the influences of uncertainties and nonlinearities in real flight environment are simulated by the formulation of average aerodynamic modeling errors （AAME）. Then the nonlinear six degrees-of-freedom （IX）F） flight is emulated on the winged aircraft model, and the results demonstrate a good tracking performance and robustness of the designed flight controller.