近年来,欠驱动水面船(以下简称欠驱动船或USV)区域保持控制一直是备受关注的研究热点和难点问题。区域保持需要对船舶的输出位置进行严格约束,保证船舶始终在规定的作业区域内运动。目前具备定点定位或区域保持功能的船舶基本上是全驱动动力定位船或大型海洋工作平台。此类船舶可以通过锚泊定位或者动力定位系统实现定点定位或区域保持控制[1-2]。然而,目前海上航行的大多数船舶的独立控制输入量少于自由度数量(横向缺少推进器),属于带有加速度约束的二阶非完整系统,即具有欠驱动特性[3-4]。由于其欠驱动特性,船舶无法提供横向推力,无法对其进行定点定位控制;并且多数欠驱动船结构简单体型小也无法采用锚泊定位或动力定位系统实现区域保持控制[5-6]。目前对欠驱动船区域保持控制的研究也是寥寥无几,鲜有相关参考文献。因此,对于欠驱动船区域保持控制的研究非常具有挑战性和实际意义。
一方面,通过已有文献拟考虑将欠驱动船区域保持问题转化为输出状态变量约束系统的控制问题[7]。当欠驱动船执行区域边界具有严格限制的相关任务时(如区域内待命、水文勘测、敌情侦查等)[8],为确保欠驱动船低速区域保持的高性能和安全性,通常对系统的输出状态变量做出严格限制。此外,小型欠驱动船在实际海洋环境中进行航行与作业时,易受风浪流等外界环境干扰影响。因此,不依赖精确数学模型和对外界环境干扰具有良好鲁棒性的滑模控制技术受到广泛研究和应用[9-10]。
另一方面,为了解决输出状态变量约束问题,研究人员提出了许多方法,如模型预测控制[11]、定值调节器[12]等。但是上述控制方法本质上均是基于数值计算的,会加剧计算过程,容易引起计算量过大。为了避免上述问题,一种基于障碍李雅普诺夫函数(Barrier Lyapunov function, BLF)的方法被提出并用于处理输出约束问题。它的基本思想是当自变量的值趋近于规定区域边界时,BLF的值将趋近于无穷大,因此通过BLF的有界性,可以达到限制输出的目的。该方法最早被应用于解决具有布鲁诺夫斯基形式(Brunovsky)系统[13]的输出约束问题。经过发展,文献[14-15]针对多状态约束下全驱动水面船的轨迹跟踪问题,采用对称障碍李雅普诺夫函数设计轨迹跟踪控制器解决输出约束问题,有效保证系统受约束的输出变量不超出约束边界。文献[16]对一类输出约束的严格反馈单输入单输出系统,分别基于BLF和ABLF(非对称障碍李雅普诺夫函数)提出了2种鲁棒自适应控制方案。但是上述方法均应用于全驱动系统并未考虑欠驱动系统的情况。
综合上述讨论,本文的主要贡献在于首先控制器设计不依赖精确数学模型,考虑时变外界环境干扰;然后,在运动学回路中设计期望速度并将其视为虚拟输入,在动力学回路中利用鲁棒滑模方法实现对参考速度的跟踪控制;最后将区域边界限制转换为输出位置状态约束解决了欠驱动船的区域保持问题。
1 问题描述及建模 1.1 预备知识定义1[16] 对于D是包含原点的开域,障碍李雅普诺夫函数V(x)是定义在D上的关于系统ẋ=f(t, x)的标量函数,满足光滑正定;在D中的每一点有连续的一阶偏导数;当x趋近D的边界时,V(x)→∞;当x(0)∈D时,有V(x)≤b, ∀t≥0,其中b>0。
引理1[16] 对任意常数kb>0,假设S1:={s1∈R:|s1| < kb}⊂R,N:=Rl×S1⊂Rl+1是开集。对于系统ẋ=f(t, x),其中x:=[ω s1]T∈N是系统的状态,函数f:=R+×N→Rl+1关于变量t分段连续且在R+×N, S1满足Lipschitz条件。假设存在函数U:=Rl→R+和V1:=S1→R+,在各自区域内是连续可导且正定,那么s1→-kb或s1→kb时,有V1(s1)→∞,γ1(‖ω‖)≤U(ω)≤γ2(‖ω‖),其中γ1和γ2是κ∞函数。假设V(x):=V1(s1)+U(ω),s1(0)∈s1∈(-kb, kb),如果满足不等式:
$ \dot V = (\partial V/\partial x)f \le 0 $ |
成立,那么有如下结论成立:对∀t∈[0, ∞),s1(t)仍在集合s1∈(-kb, kb)中。
1.2 欠驱动水面船模型考虑外界环境干扰下USV水平面运动学和动力学模型为[17](忽略横摇、纵摇和垂荡):
$ \left\{ {\begin{array}{*{20}{l}} {\dot x = u{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - v{\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }\\ {\dot y = u{\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi + v{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }\\ {\dot \psi = r} \end{array}} \right. $ | (1) |
$ \left\{ {\begin{array}{*{20}{l}} {\dot u = (1/{m_u})[{m_v}vr - {d_u}u + {\tau _u} + {\tau _{wu}}(t)]}\\ {\dot v = (1/{m_v})[ - {m_u}ur - {d_v}v + {\tau _{wv}}(t)]}\\ {\dot r = (1/{m_r})[({m_u} - {m_v})uv - {d_r}r + {\tau _r} + {\tau _{wr}}(t)]} \end{array}} \right. $ | (2) |
式中:x、y、ψ表示北东坐标系下USV位置坐标和艏向角;u、v、r表示船体坐标系下纵向、横向速度和艏向回转速度;mi, di, i=u, v, r表示欠驱动船系统包括附加质量在内的惯性参数和水动力阻尼系数,为已知项;τu和τr表示纵向推力和转艏力矩,由于没有横向控制输入,因此所研究的为欠驱动船区域保持的控制问题。τwi表示由风浪流等引起的外界环境干扰力和力矩。
在接下来的控制器设计中,定义区域保持误差变量为de=
![]() |
Download:
|
图 1 欠驱动船区域保持结构框图 Fig. 1 Structural block diagram of area-keeping for USV |
假设1 外界环境干扰有界,满足|τwi|≤τwimax < ∞,0 < τwimax < ∞是未知常数。
假设2 欠驱动船重心位于坐标原点,USV左右对称,且非对角元素相对于对角元素为小量可忽略。
假设3 存在最小航速umin和最大航速umax,使得期望纵向速度满足αu∈[umin, umax], umax>umin>0。
假设4 期望航迹光滑有界Ωd:[0, ∞]→R2,且对时间的导数均有界。
假设5 所有状态都是可测量的,并可供反馈。
2 控制器设计和稳定分析 2.1 基于对称障碍李雅普诺夫函数的鲁棒滑模控制器设计首先定义欠驱动船的位置和速度跟踪误差变量xe、ye、ue、ve,两边求导并代入式(1)和(2)可得:
$ \left\{ \begin{array}{l} {{\dot x}_e} = \dot x - {{\dot x}_d} = u{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - v{\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - {{\dot x}_d}\\ {{\dot y}_e} = \dot y - {{\dot y}_d} = u{\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi + v{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - {{\dot y}_d}\\ {{\dot u}_e} = \dot u - {{\dot \alpha }_u} = (1/{m_u})[{m_v}vr - {d_u}u + \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\tau _u} + {\tau _{wu}}(t)] - {{\dot \alpha }_u}\\ {{\dot v}_e} = \dot v - {{\dot \alpha }_v} = (1/{m_v})[ - {m_u}ur - {d_v}v + \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\tau _{wv}}(t)] - {{\dot \alpha }_v} \end{array} \right. $ | (3) |
为避免如注1所示虚拟控制律的设计缺陷,同时实现位置跟踪误差收敛,根据文献[9]设计期望纵向和横向速度即虚拟控制律如下,
$ \begin{array}{l} \left[ {\begin{array}{*{20}{l}} {{\alpha _u}}\\ {{\alpha _v}} \end{array}} \right] = \left[ {\begin{array}{*{20}{c}} {{\rm{cos}}\psi }&{{\rm{sin}}\psi }\\ { - {\rm{sin}}\psi }&{{\rm{cos}}\psi } \end{array}} \right] \cdot \\ \left[ {\begin{array}{*{20}{c}} {{{\dot x}_d} + {h_1}{\rm{tanh}}( - {k_1}{x_e}/{h_1})}\\ {{{\dot y}_d} + {h_2}{\rm{tanh}}( - {k_2}{y_e}/{h_2})} \end{array}} \right] \end{array} $ | (4) |
式中:xd、yd和αu、αv分别表示期望位置和期望速度,k1, k2>0和h1, h2>0表示待设计参数,h1、h2根据欠驱动船最大速度的物理限制选取,而k1、k2可以影响欠驱动船的收敛速度。
同理可知,
$ \left[ {\begin{array}{*{20}{c}} {{u_e}}\\ {{v_e}} \end{array}} \right] = \mathit{\boldsymbol{J}}(\psi )\left[ {\begin{array}{*{20}{c}} {{{\dot x}_e} - {h_1}{\rm{tanh}}( - {k_1}{x_e}/{h_1})}\\ {{{\dot y}_e} - {h_2}{\rm{tanh}}( - {k_2}{y_e}/{h_2})} \end{array}} \right] $ | (5) |
式中:
$ \begin{array}{*{20}{l}} {{{\dot x}_e} = {h_1}{\rm{tanh}}( - {k_1}{x_e}/{h_1})}\\ {{{\dot y}_e} = {h_2}{\rm{tanh}}( - {k_2}{y_e}/{h_2})} \end{array} $ | (6) |
接下来证明xe、ye的收敛性,选取如下李雅普诺夫函数V1为:
$ {V_1} = \frac{1}{2}x_e^2 + \frac{1}{2}y_e^2 $ | (7) |
将V1对时间求导并代入式(6)可得:
$ \begin{array}{*{20}{c}} {{{\dot V}_1} = {x_e}{{\dot x}_e} + {y_e}{{\dot y}_e} = - {h_1}{x_e}{\rm{tanh}}({k_1}{x_e}/{h_1}) - }\\ {{h_2}{y_e}{\rm{tanh}}({k_2}{y_e}/{h_2})} \end{array} $ | (8) |
已知k1, k2>0和h1, h2>0,显然对∀(xe, ye)≠(0, 0),满足
然后,为使欠驱动船跟踪期望速度αu和αv,设计非线性滑模面:
$ {{L_1} = {u_e} + {\xi _1}\int_0^t {{u_e}} (\tau ){\rm{d}}\tau + {\xi _2}} $ | (9) |
$ {{L_2} = {{\dot v}_e} + {\xi _3}\int_0^t {{v_e}} (\tau ){\rm{d}}\tau + {\xi _4}{v_e}} $ | (10) |
式中:待设计参数ξ1, ξ2, ξ3, ξ4>0。对滑模面求导,并将式(2)和(5)代入得:
$ \begin{array}{*{20}{c}} {{{\dot L}_1} = (1/{m_u})[{m_v}vr - {d_u}u + {\tau _u} + {\tau _{wu}}(t)] - }\\ {{{\dot \alpha }_u} + {\xi _1}{u_e}} \end{array} $ | (11) |
$ \begin{array}{l} {\kern 1pt} {\kern 1pt} \begin{array}{*{20}{c}} {{{\dot L}_2} = {{\ddot v}_e} + {\xi _3}{v_e} + {\xi _4}{{\dot v}_e} = }\\ {(1/{m_v})[ - {m_u}(\dot ur + u\dot r) - {d_v}\dot v + {{\dot \tau }_{wv}}(t)] - } \end{array}\\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {{\ddot \alpha }_v} + {\xi _3}{v_e} + {\xi _4}{{\dot v}_e} = \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} (1/{m_v}) \cdot \\ \left[ { - {m_u}(\dot ur + (u/{m_r})(({m_u} - {m_v})uv - {d_r}r + } \right.\\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\tau _r} + {\tau _{wr}}(t))) - {d_v}\dot v + {{\dot \tau }_{wv}}(t)] - \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {{\ddot \alpha }_v} + {\xi _3}{v_e} + {\xi _4}{{\dot v}_e} \end{array} $ | (12) |
分别令
为了满足可达性保证系统能在有限时间内抵达滑模面,需进一步设计切换控制律τu2和τr2使滑模面具有:
$ {{{\dot L}_1} = - {C_1} {\rm{sign}} ({L_1})} $ | (13) |
$ {{{\dot L}_2} = - {C_2} {\rm{sign}} ({L_2})} $ | (14) |
式中C1, C2>0表示待设计控制参数。
最后,为了便于分析,本文假设作业区域是以O(0, 0)为圆心,R为半径的圆形区域,而期望航迹半径为ra且ra < R。则区域保持任务即可描述为始终保持欠驱动船的运动轨迹在圆内,不超出规定区域边界。
考虑外界环境干扰和位置状态变量约束条件,采用对称障碍李雅普诺夫函数设计如下纵向推力和转艏力矩,有:
$ \begin{array}{*{20}{l}} {{\tau _{u1}} = - {m_v}vr + {d_u}u - {\tau _{wu}}(t) + {m_u}{{\dot \alpha }_u} - }\\ {{\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {m_u}{\xi _1}{u_e} - ({m_u}/{L_1})\frac{{{x_e}{{\dot x}_e} + {y_e}{{\dot y}_e}}}{{k_d^2 - d_e^2}}} \end{array} $ | (15) |
$ \begin{array}{*{20}{c}} {{\tau _{r1}} = \frac{{{m_r}}}{{u{m_u}}}( - {d_v}\dot v + {{\dot \tau }_{wv}}(t)) + \frac{{{m_v}{m_r}}}{{u{m_u}}}(\chi - {\alpha _u}\dot r - }\\ {{\xi _3}{v_e} - {\xi _4}{{\dot v}_e}) + ({m_v} - {m_u})uv + {d_r}r - {\tau _{wr}}(t) - \frac{{{m_r}}}{u}\dot ur} \end{array} $ | (16) |
选取切换控制律τu2和τr2:
$ {{\tau _{u2}} = - {m_u}{C_1} {\rm{sign}} ({L_1})} $ | (17) |
$ {{\tau _{r2}} = \frac{{{m_v}{m_r}}}{{u{m_u}}}{C_2} {\rm{sign}} ({L_2})} $ | (18) |
可得:
$ {{\tau _u} = {\tau _{u1}} + {\tau _{u2}}} $ | (19) |
$ {{\tau _r} = {\tau _{r1}} + {\tau _{r2}}} $ | (20) |
中间变量形式为:
$ \begin{gathered} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} \chi = - {{\dddot x}_d}{\text{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi + {{\dddot y}_d}{\text{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - {{\ddot x}_d}r{\text{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - \hfill \\ {{\ddot y}_d}r{\text{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - {{\dot \alpha }_u}r + {\gamma _1}r{\text{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi + {\gamma _2}r{\text{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi + \hfill \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {{\dot \gamma }_1}{\text{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi - {{\dot \gamma }_2}{\text{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi \hfill \\ \end{gathered} $ |
$ {{\gamma _1} = {k_1}{{\dot x}_e} {\rm{sech}}{ ^2}( - {k_1}{x_e}/{h_1})} $ |
$ {{\gamma _2} = {k_2}{{\dot y}_e} {\rm{sech}}{ ^2}( - {k_2}{y_e}/{h_2})} $ |
注1 虚拟控制律选用式(4),而不选取:
$ \left[ {\begin{array}{*{20}{l}} {{\alpha _u}}\\ {{\alpha _v}} \end{array}} \right] = \left[ {\begin{array}{*{20}{c}} {{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{{\rm{sin}}{\kern 1pt} {\kern 1pt} \psi }\\ { - {\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi } \end{array}} \right]\left[ {\begin{array}{*{20}{c}} {{{\dot x}_d} - k{x_e}}\\ {{{\dot y}_d} - k{y_e}} \end{array}} \right] $ |
是考虑当|xe|和|ye|较大时,会导致虚拟控制律超出欠驱动船所能达到的最大速度,从而无法确保设计的控制输入τu与τr使得USV的纵向和横向速度跟踪期望速度αu和αv。
注2 由于L1≠0,所以当kd=de时,控制律τu(t)将产生奇异值,而区域保持控制器设计过程需满足de < kd;当u=0时,控制律τr(t)将产生奇异值,而在假设3条件中纵向速度不会为零;综上所提区域保持控制器不会出现奇异值情况。
2.2 稳定性分析定理1 针对区域保持滑模控制律(19)和(20)作用下的USV水平面运动闭环系统(1)(2)和跟踪误差(3),通过选取合适的参数k1、k2、h1、h2、ξ1、ξ2、ξ3、ξ4,所提控制律不仅能使位置误差xe、ye和速度跟踪误差ue、ve渐近收敛于任意小零值邻域内,且能保证USV回转速度r的有界性,并且当区域保持误差初始值满足de(0)=
证明 首先,在式(3)和式(6)下对式(4)求一阶和二阶导数,为:
$ \begin{array}{l} {\kern 1pt} {\kern 1pt} \left[ {\begin{array}{*{20}{c}} {{{\dot \alpha }_u}}\\ {{{\dot \alpha }_v}} \end{array}} \right] = r\left[ {\begin{array}{*{20}{c}} { - {\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }\\ { - {\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{ - {\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi } \end{array}} \right] \cdot \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} \left[ {\begin{array}{*{20}{c}} {{{\dot x}_d} + {h_1}{\rm{tanh}}( - {k_1}{x_e}/{h_1})}\\ {{{\dot y}_d} + {h_2}{\rm{tanh}}( - {k_2}{y_e}/{h_2})} \end{array}} \right] + \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} \left[ {\begin{array}{*{20}{c}} {{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{{\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }\\ { - {\rm{sin}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi }&{{\rm{cos}}{\kern 1pt} {\kern 1pt} {\kern 1pt} \psi } \end{array}} \right] \cdot \\ \left[ {\begin{array}{*{20}{l}} {{{\ddot x}_d} - {k_1}{{\dot x}_e} {\rm{sech}}{ ^2}( - {k_1}{x_e}/{h_1})}\\ {{{\ddot y}_d} - {k_2}{{\dot y}_e} {\rm{sech}}{ ^2}( - {k_2}{y_e}/{h_2})} \end{array}} \right] \end{array} $ | (21) |
$ {{\ddot \alpha }_v} = \chi - {\alpha _u}\dot r $ | (22) |
式中χ为控制律(19)和(20)的中间变量。
然后,为了使约束变量de始终保持在约束区域内,选取带有障碍李雅普诺夫函数的Lyapunov函数V2:
$ {V_2} = \frac{1}{2}L_1^2 + \frac{1}{2}L_2^2 + \frac{1}{2}{\rm{ln}}\frac{{k_d^2}}{{k_d^2 - d_e^2}} $ | (23) |
将V2两边求导,并代入式(11)和(12)可得,
$ \begin{array}{l} {\kern 1pt} \begin{array}{*{20}{c}} {{{\dot V}_2} = {L_1}{{\dot L}_1} + {L_2}{{\dot L}_2} + \frac{{{d_e}{{\dot d}_e}}}{{k_d^2 - d_e^2}} = }\\ {{L_1}\left[ {\frac{1}{{{m_u}}}({m_v}vr - {d_u}u + {\tau _u} + {\tau _{wu}}(t)) - {{\dot \alpha }_u} + {\xi _1}{u_e}} \right] + } \end{array}\\ {L_2}\left\{ {\frac{1}{{{m_v}}}\left[ { - {m_u}(\dot ur + \frac{u}{{{m_r}}}(({m_u} - {m_v})uv - {d_r}r + {\tau _r} + } \right.} \right.\\ \left. {{\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\tau _{ur}}(t)) - {d_v}\dot v + {{\dot \tau }_{wv}}(t)] - {{\ddot \alpha }_v} + {\xi _3}{v_e} + {\xi _4}{{\dot v}_e}} \right\} + \\ {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} \frac{{{x_e}{{\dot x}_e} + {y_e}{{\dot y}_e}}}{{k_d^2 - d_e^2}} \end{array} $ | (24) |
由式(24)且根据引理1可知,当误差信号de的初始值满足|de(0)| < kd,则对∀t>0有|de(t)| < kd恒成立。
代入式(19)和(20)可得,
$ \begin{array}{*{20}{l}} {{{\dot V}_2} = - {C_1}{L_1} {\rm{sign}} ({L_1}) - {C_2}{L_2} {\rm{sign}} ({L_2}) = }\\ {{\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} {\kern 1pt} - {C_1}|{L_1}| - {C_2}|{L_2}|} \end{array} $ | (25) |
根据C1, C2>0,所以对∀(L1, L2)≠(0, 0),
$ {{\dot u}_e} = - {\xi _1}{u_e} $ | (26) |
$ {{\ddot v}_e} = - {\xi _3}{v_e} - {\xi _4}{{\dot v}_e} $ | (27) |
由于参数ξ1, ξ3, ξ4>0和式(21)、(22),所以动态特性(26)和(27)保证了速度误差ue和ve能够渐近收敛于零值邻域。因此根据式(4)和(6)xe, ye也渐近收敛于零值邻域内。
最后,证明艏向回转速度r的有界性,选取李雅普诺夫函数V3为:
$ {V_3} = \frac{1}{2}{m_r}{r^2} $ | (28) |
将V3两边求导并代入式(2)有:
$ {\dot V_3} = {m_r}r\dot r = r[({m_u} - {m_v})uv + {\tau _r} + {\tau _{wr}}(t) - {d_r}r] $ | (29) |
若满足如下条件,则
$ \begin{array}{*{20}{l}} {r > 0,{d_r}r > ({m_u} - {m_v})uv + {\tau _r} + {\tau _{ur}}(t)}\\ {r < 0,{d_r}r < ({m_u} - {m_v})uv + {\tau _r} + {\tau _{wr}}(t)} \end{array} $ | (30) |
已知dr>0,式(30)可重写为:
$ |{d_r}r| > \left| {({m_u} - {m_v})uv + {\tau _r} + {\tau _{wr}}(t)} \right| $ | (31) |
当满足条件(31)时,
综上,根据稳定性分析可知欠驱动船的位置误差xe, ye是有界的渐近收敛于零值邻域内,纵向和横向速度分别渐近收敛于期望纵向和横向速度,同时艏向回转速度r有界。此外,当区域保持误差初始值满足de(0) < kd条件时,对∀t>0,有|de(t)| < kd恒成立,USV区域保持闭环系统是半全局渐近稳定的。综上定理1得证。
3 仿真实验采用文献[18]中CyberShipⅡ数学模型(1:70)进行仿真研究,针对风浪流外界环境干扰下欠驱动船区域保持控制问题,结合对称障碍Lyapunov函数将区域保持问题转换为位置状态变量约束问题,并利用鲁棒滑模控制器进行数值仿真,分别在有无外界环境干扰情况下进行对比仿真,验证所提USV区域保持控制器的有效性和鲁棒性。为了不失一般性,对比仿真中的区域保持控制器均设置相同的控制参数。
其中CyberShipⅡ模型参数如下,船长1.255 m,质量23.8 kg,其他参数为,
$ \begin{array}{*{20}{l}} {{\tau _{wu}} = 0.05 \times [1 + 0.03{\rm{cos}}(0.3t) + 0.05{\rm{sin}}(0.5t) + }\\ {0.01{\rm{sin}}(0.1t)]{\tau _{wv}} = 0.5 \times [1 + 0.03{\rm{cos}}(0.3t) + }\\ {0.05{\rm{sin}}(0.5t) + 0.01{\rm{sin}}(0.1t)]{\tau _{wr}} = 5 \times [1 + }\\ {0.03{\rm{cos}}(0.3t) + 0.05{\rm{sin}}(0.5t) + 0.01{\rm{sin}}(0.1t)]} \end{array} $ |
![]() |
Download:
|
图 2 欠驱动水面船区域保持位置变化曲线 Fig. 2 The position curve of area keeping for USV |
![]() |
Download:
|
图 3 欠驱动水面船区域保持位置误差约束曲线 Fig. 3 position error of area keeping for USV |
![]() |
Download:
|
图 4 欠驱动水面船区域保持速度误差曲线 Fig. 4 The velocity error of area keeping for USV |
![]() |
Download:
|
图 5 欠驱动水面船区域保持的艏向回转速度变化曲线 Fig. 5 The yaw angular velocity of area keeping for USV |
![]() |
Download:
|
图 6 欠驱动水面船区域保持控制输出曲线 Fig. 6 The control force outputs of area keeping for USV |
从图 2欠驱动船区域保持位置变化曲线可以看出,根据位置状态约束条件选取对称障碍李雅普诺夫函数设计的滑模控制律,当欠驱动船初始位置在约束区域内时,不论在有无外界环境干扰下USV位置状态始终满足约束条件,即欠驱动船始终在规定区域内运动,不会超出约束边界,可以得出所设计控制器是有效的同时也证明了所提控制器对外界环境干扰具有一定的鲁棒性。由图 3欠驱动船区域保持位置误差约束曲线显示位置误差是有界收敛的,并且在区域保持允许的误差范围内。从图 4欠驱动水面船区域保持的速度误差变化曲线可以看出,相比较于无外界环境干扰的情况,存在外界环境干扰时速度误差波动较大,但2种情况下速度误差均可较快的收敛于零。
通过图 5的艏向回转速度变化曲线可以明显看出,相比于无外界环境干扰情况,为适应外界环境干扰艏向回转速度变化较大。而对比图 6的欠驱动船区域保持控制输入变化曲线可知,存在外界环境干扰时欠驱动船提供的纵向推力和转艏力矩相比于无干扰下也有一定的波动,以抵抗风浪流干扰。
4 结论1) 本文将风浪流外界环境干扰下欠驱动水面船的区域保持控制问题转换为位置状态变量约束问题,提出了一种基于对称障碍李雅普诺夫函数的鲁棒滑模控制方法,实现位置状态变量约束控制保证欠驱动船不超出约束边界。
2) 所提方法通过反步法将区域保持控制器分解为运动学和动力学回路,在运动学回路中设计期望速度并将其视为虚拟控制输入,在动力学回路中利用滑模方法实现对参考速度的跟踪控制。
3) 此外,由于区域保持控制对USV在区域内的运动方式和收敛速度不作要求,因此在本文基础上考虑事件触发思想达到减少控制器计算量和执行机构损耗将是后续研究的重点。
[1] |
SØRENSEN A J. A survey of dynamic positioning control systems[J]. Annual reviews in control, 2011, 35(1): 123-136. DOI:10.1016/j.arcontrol.2011.03.008 ( ![]() |
[2] |
TUO Yulong, WANG Yuanhui, WANG Shasha. Reliability-based robust online constructive fuzzy positioning control of a turret-moored floating production storage and offloading vessel[J]. IEEE access, 2018, 6: 36019-36030. DOI:10.1109/ACCESS.2018.2851286 ( ![]() |
[3] |
LIU Zhixiang, ZHANG Youmin, YU Xiang, et al. Unmanned surface vehicles:an overview of developments and challenges[J]. Annual reviews in control, 2016, 41: 71-93. DOI:10.1016/j.arcontrol.2016.04.018 ( ![]() |
[4] |
付明玉, 王莎莎, 王元慧, 等. 带输入饱和的欠驱动水面船参数自适应滑模控制[J]. 哈尔滨工程大学学报, 2019, 40(1): 202-209. FU Mingyu, WANG Shasha, WANG Yuanhui, et al. Parameter adaptive sliding mode control for underactuated surface vehicle with input saturation[J]. Journal of Harbin Engineering University, 2019, 40(1): 202-209. ( ![]() |
[5] |
肖本华.动力定位自动区域保持控制方法研究[D].哈尔滨: 哈尔滨工程大学, 2002: 60-75. XIAO Benhua. Research on control algorithm of auto area keeping for dynamic positioning system[D]. Harbin: Harbin Engineering University, 2002: 60-75. ( ![]() |
[6] |
邹金锟.基于运动仿真的无人艇区域保持方法研究[D].哈尔滨: 哈尔滨工程大学, 2014: 52-71. ZOU Jinkun. Research on area keeping method of USV based on motion simulation[D]. Harbin: Harbin Engineering University, 2014: 52-71. ( ![]() |
[7] |
TEE K P, REN Beibei, GE S S. Control of nonlinear systems with time-varying output constraints[J]. Automatica, 2011, 47(11): 2511-2516. DOI:10.1016/j.automatica.2011.08.044 ( ![]() |
[8] |
CAMPBELL S, NAEEM W, IRWIN G W. A review on improving the autonomy of unmanned surface vehicles through intelligent collision avoidance manoeuvres[J]. Annual reviews in control, 2012, 36(2): 267-283. DOI:10.1016/j.arcontrol.2012.09.008 ( ![]() |
[9] |
朱齐丹, 于瑞亭, 夏桂华, 等. 风浪流干扰及参数不确定欠驱动船舶航迹跟踪的滑模鲁棒控制[J]. 控制理论与应用, 2012, 29(7): 959-964. ZHU Qidan, YU Ruiting, XIA Guihua, et al. Sliding-mode robust tracking control for underactuated surface vessels with parameter uncertainties and external disturbances[J]. Control theory & applications, 2012, 29(7): 959-964. ( ![]() |
[10] |
ELMOKADEM T, ZRIBI M, YOUCEF-TOUMI K. Terminal sliding mode control for the trajectory tracking of underactuated autonomous underwater vehicles[J]. Ocean engineering, 2016, 129: 613-625. ( ![]() |
[11] |
MAYNE D Q, RAWLINGS J B, RAO C V, et al. Constrained model predictive control:stability and optimality[J]. Automatica, 2000, 36(6): 789-814. DOI:10.1016/S0005-1098(99)00214-9 ( ![]() |
[12] |
BEMPORAD A. Reference governor for constrained nonlinear systems[J]. IEEE transactions on automatic control, 1998, 43(3): 415-419. ( ![]() |
[13] |
NGO K B, MAHONY R, JIANG Zhongping. Integrator backstepping using barrier functions for systems with multiple state constraints[C]//Proceedings of the 44th IEEE Conference on Decision and Control. Seville, Spain, Spain, 2005: 8306-8312.
( ![]() |
[14] |
ZHAO Zhen, HE Wei, GE S S. Adaptive neural network control of a fully actuated marine surface vessel with multiple output constraints[J]. IEEE transactions on control systems technology, 2014, 22(4): 1536-1543. ( ![]() |
[15] |
HE Wei, YIN Zhao, SUN Changyin. Adaptive neural network control of a marine vessel with constraints using the asymmetric barrier lyapunov function[J]. IEEE transactions on cybernetics, 2017, 47(7): 1641-1651. DOI:10.1109/TCYB.2016.2554621 ( ![]() |
[16] |
TEE K P, GE S S, TAY E H. Barrier lyapunov functions for the control of output-constrained nonlinear systems[J]. Automatica, 2009, 45(4): 918-927. DOI:10.1016/j.automatica.2008.11.017 ( ![]() |
[17] |
LIAO Yulei, ZHANG Minjun, WAN Lei, et al. Trajectory tracking control for underactuated unmanned surface vehicles with dynamic uncertainties[J]. Journal of Central South University, 2016, 23(2): 370-378. DOI:10.1007/s11771-016-3082-4 ( ![]() |
[18] |
SKJETNE R, SMOGELI Ø N, FOSSEN T I. A nonlinear ship manoeuvering model:identification and adaptive control with experiments for a model ship[J]. Modeling, identification and control, 2004, 25(1): 3-27. ( ![]() |