基于非线性模型预测的四旋翼无人机轨迹跟踪控制
Trajectory Tracking Control of Quadrotor Based on Nonlinear Model Predictive Control
闫玉盼 1饶兵 2刘砚菊 1冯迎宾1
作者信息
- 1. 沈阳理工大学 自动化与电气工程学院,沈阳 110159
- 2. 沈阳天眼智云智能技术研究院有限公司,沈阳 110179
- 折叠
摘要
针对四旋翼无人机轨迹跟踪控制器跟踪性能不稳定的问题,提出一种基于非线性模型预测控制(nolinear model predict control,NMPC)的轨迹跟踪算法.首先,建立四旋翼无人机的动力学模型,定义四旋翼无人机的位置和姿态为状态量,螺旋桨转速为控制输入量,建立非线性状态空间方程作为控制算法的预测模型.其次,定义最优化函数和四旋翼无人机控制约束,将轨迹跟踪控制问题转换为非线性最优化求解问题.最后,通过多重打靶法求解得到的最优控制量作为四旋翼无人机的输入信号.为验证NMPC算法的跟踪性能,在Matlab中搭建仿真平台进行对比实验,结果表明,与PID和串级模型预测控制(model predict control,MPC)及改进MPC方法相比,NMPC算法能够在满足约束的情况下完成轨迹跟踪任务,误差小、精度高,并具有抗干扰能力.
Abstract
For the problem of unsteady tracking performance of quadrotor trajectory tracking con-troller,a trajectory tracking algorithm based on nonlinear model predictive control(NMPC)is pro-posed.Firstly,the quadrotor's dynamic model is established,position and attitude of the quadrotor are defined as state variables,the propellers'speed are defined as control input variables.The non-linear state-space equation is established as the prediction model for the control algorithm.Then,the trajectory tracking control problem is transformed into nonlinear optimization problem as the optimi-zation function and the quadrotor control constraints are defined.Finally,the optimal quantity solved by multiple shooting method are used as control input for the quadrotor.In order to verify the track-ing performance of NMPC algorithm,comparison simulation experiment is conducted in Matlab,in which a quadrotor platform is built.Experiment results show that compared with PID,cascade mod-el predictive control(MPC)and improved MPC methods,the trajectory tracking under the condition of constraints could be finished by the NMPC algorithm,with lower error,higher accuracy and anti-interference ability.
关键词
四旋翼无人机/非线性模型预测控制/轨迹跟踪Key words
quadrotor unmanned aerial vehicle/nonlinear model predict control/trajectory tracking引用本文复制引用
基金项目
辽宁省教育厅高等学校基本科研项目(LJKMZ20220614)
出版年
2024
NETL
NSTL
万方数据