基于主动建模的无人直升机增强LQR控制

为了解决无人直升机控制问题,通过把主动建模与LQR(Linear Quadratic Regulator)控制相结合,提出一种能补偿模型差的控制方法。该方法在悬停状态下,采用简化模型设计LQR控制器,并通过UKF(Un-scented-Kalman-Filter)在线估计简化模型与全状态模型的模型差,使用模型差作为补偿项对LQR控制增强。针对实际直升机动力学模型进行仿真,验证了基于UKF的估计和增强LQR控制的有效性。仿真实验结果证明,基于UKF的主动建模技术能够快速估计状态和参数变化,并且增强LQR控制能够使系统适应模型不确定性。

A control method that can compensate model error by integrating active model into LQR(Linear Quadratic Regulator) control is proposed.In the scheme,a normal LQR control designed from a simplified model at hovering is enhanced by means of UKF(Unscented-Kalman-Filter) based estimation,which tries to capture the model error between the simplified model and the full dynamics.Simulations about helicopter model are conducted to verify both the UKF-based estimation and the enhanced LQR control.Simulation results demonstrate that the UKF-based online modeling technique estimate quickly to the changes in both state and parameters,and the enhanced LQR control makes the control system adaptive to model uncertainties autonomously.

国家863计划基金资助项目(2006AA04Z206)