王勋, 张代兵, 沈林成. 一种基于虚拟力的无人机路径跟踪控制方法[J]. 机器人, 2016, 38(3): 329-336. DOI: 10.13973/j.cnki.robot.2016.0329
引用本文: 王勋, 张代兵, 沈林成. 一种基于虚拟力的无人机路径跟踪控制方法[J]. 机器人, 2016, 38(3): 329-336. DOI: 10.13973/j.cnki.robot.2016.0329
WANG Xun, ZHANG Daibing, SHEN Lincheng. A Virtual Force Based Path Following Approach for Unmanned Aerial Vehicles[J]. ROBOT, 2016, 38(3): 329-336. DOI: 10.13973/j.cnki.robot.2016.0329
Citation: WANG Xun, ZHANG Daibing, SHEN Lincheng. A Virtual Force Based Path Following Approach for Unmanned Aerial Vehicles[J]. ROBOT, 2016, 38(3): 329-336. DOI: 10.13973/j.cnki.robot.2016.0329

一种基于虚拟力的无人机路径跟踪控制方法

A Virtual Force Based Path Following Approach for Unmanned Aerial Vehicles

  • 摘要: 提出一种基于虚拟力的无人机路径跟踪控制方法.通过设计虚拟向心力、虚拟弹簧力和虚拟阻力计算期望的转向速率.其中虚拟向心力可以补偿参考路径曲率,虚拟弹簧力使无人机收敛到参考路径上,虚拟阻力能够在收敛过程中防止震荡的产生.该方法不仅可以跟踪直线和圆形路径,还可以精确跟踪变曲率曲线.在跟踪直线时,该方法等价于比例—微分控制;跟踪圆形或变曲率曲线时,等价于反馈线性化方法.论文分析了该方法的稳定性和收敛性,考虑了输入约束对该方法跟踪性能的影响.利用虚拟力控制无人机,使控制参数具有明确的物理意义,从而使参数在实际应用中容易整定.仿真结果证明该方法是有效的,且跟踪性能优于 NLGL(非线性导航逻辑)方法.

     

    Abstract: A virtual force based path following approach is presented for unmanned aerial vehicles (UAVs). Three virtual forces, including a virtual centripetal force, a virtual spring force and a virtual drag force, are designed to calculate the desired heading rate. The virtual centripetal force compensates the influence of the reference curvature. The virtual spring force ensures the vehicle converge to the reference path. Meanwhile, the virtual drag force prevents oscillation in the convergence process. The proposed approach can be used to accurately follow straight-line, circle, as well as curve with time-varying curvature. The approach is equivalent to a proportional-derivative controller when following a straight-line, equivalent to a feedback linearization method when following circular or curved paths. The stability and convergence are analyzed. The influence of the input constraint on the following performance is considered. When using virtual forces to control UAVs, the physical meanings of the parameters are definite, which makes them easy to tune in application. Simulation results demonstrate the effectiveness of the proposed approach, and its performance is better than the NLGL (nonlinear guidance logic) approach.

     

/

返回文章
返回