线性系统与非线性系统的区别
我们在读论文的时候经常会遇到这两个系统,线性系统与非线性系统,这两者之间有什么区别呢?
线性指量与量之间按比例、成直线的关系,在空间和时间上代表规则和光滑的运动;非线性则指不按比例、不成直线的关系代表不规则的运动和突变。
在判断是否是线性系统上,主要看叠加原理(Superposition)!!!
系统的方程为 x ˙ = f ( x ) \dot x = f(x) x˙=f(x) ,如果 x 1 , x 2 x_1,x_2 x1,x2 是方程的解,有 x 3 = k 1 x 1 + k 2 x 2 ( k 1 , k 2 ∈ R ) x_3 = k_1 x_1 + k_2 x_2(k_1,k_2 \in \R) x3=k1x1+k2x2(k1,k2∈R) ,且 x 3 x_3 x3 也是方程的解,则系统符合叠加原理,为线性系统。
举个例子:
x
¨
+
2
x
˙
+
2
x
=
0
(
线性系统
)
x
¨
+
2
x
˙
+
2
x
2
=
0
(
非线性系统
)
x
¨
+
2
s
i
n
(
x
˙
)
+
2
x
=
0
(
非线性系统
)
\ddot x + 2 \dot x + \sqrt{2} x = 0 (线性系统) \\ \ddot x + 2 \dot x + \sqrt{2} x ^ 2 = 0 (非线性系统) \\ \ddot x + 2 sin(\dot x) + \sqrt{2} x = 0 (非线性系统) \\
x¨+2x˙+2x=0(线性系统)x¨+2x˙+2x2=0(非线性系统)x¨+2sin(x˙)+2x=0(非线性系统)
线性化方法
泰勒级数(Taylor Series)
f
(
x
)
=
f
(
x
0
)
+
f
′
(
x
0
)
1
!
(
x
−
x
0
)
+
f
′
′
(
x
0
)
2
!
(
x
−
x
0
)
2
+
.
.
.
+
f
n
(
x
0
)
n
!
(
x
−
x
0
)
n
f(x) = f(x_0) + \frac{f'(x_0)}{1 !}(x-x_0) + \frac{f''(x_0)}{2 !}(x-x_0)^2 + ...+\frac{f^{n}(x_0)}{n !}(x-x_0)^n
f(x)=f(x0)+1!f′(x0)(x−x0)+2!f′′(x0)(x−x0)2+...+n!fn(x0)(x−x0)n
如果
x
−
x
0
→
0
x-x_0 \to 0
x−x0→0,则
(
x
−
x
0
)
2
→
0
(x-x_0)^2 \to 0
(x−x0)2→0 ,则
(
x
−
x
0
)
n
→
0
(x-x_0)^n \to 0
(x−x0)n→0 。
f
(
x
)
=
f
(
x
0
)
+
f
′
(
x
0
)
(
x
−
x
0
)
=
k
2
x
+
b
f(x) = f(x_0) + f'(x_0)(x-x_0)=k_2x+b
f(x)=f(x0)+f′(x0)(x−x0)=k2x+b,其中
k
2
=
f
′
(
x
0
)
,
b
=
f
(
x
0
)
−
f
′
(
x
0
)
x
0
k_2 = f'(x_0),b = f(x_0) - f'(x_0)x_0
k2=f′(x0),b=f(x0)−f′(x0)x0 。
线性化是在某一点附近的线性化,而不是全局的线性化。
一维系统,举个例子
x ¨ + x ˙ + 1 x = 1 \ddot x + \dot x + \frac{1}{x} = 1 x¨+x˙+x1=1
在平衡点(Fixed Point)附近线性化。
x
¨
=
x
˙
=
0
⇒
1
x
=
1
⇒
x
0
=
1
\ddot x = \dot x = 0 \Rightarrow \frac{1}{x} = 1 \Rightarrow x_0 = 1
x¨=x˙=0⇒x1=1⇒x0=1
所以平衡点为
x
0
=
1
x_0 = 1
x0=1 。在
x
0
x_0
x0 附近
x
δ
=
x
0
+
x
d
x_{\delta} = x_0 + x_d
xδ=x0+xd , 所以有
x
¨
δ
+
x
˙
δ
+
1
x
δ
=
1
\ddot x_{\delta} + \dot x_{\delta} + \frac{1}{x_{\delta} } = 1
x¨δ+x˙δ+xδ1=1
运用上面的泰勒级数,先线性化
1
x
δ
\frac{1}{x_\delta}
xδ1
f
(
x
δ
)
=
f
(
x
0
)
+
f
′
(
x
0
)
(
x
δ
−
x
0
)
f(x_{\delta}) = f(x_{0}) + f'(x_0)(x_{\delta}-x_0)
f(xδ)=f(x0)+f′(x0)(xδ−x0)
1 x δ = 1 x 0 + ( − 1 x 0 2 ) ( x δ − x 0 ) = 1 x 0 − x d x 0 2 = 1 − x d \frac{1}{x_\delta} = \frac{1}{x_0} + (-\frac{1}{x_0^2})(x_\delta - x_0) = \frac{1}{x_0} -\frac{x_d}{x_0^2} = 1-x_d xδ1=x01+(−x021)(xδ−x0)=x01−x02xd=1−xd
{ x ¨ δ = x ¨ 0 + x ¨ d x ˙ δ = x ˙ 0 + x ˙ d 1 x δ = 1 − x d ⇒ x ¨ 0 + x ¨ d + x ˙ 0 + x ˙ d + 1 − x d = 1 ⇒ x ¨ d + x ˙ d − x d = 0 \begin{cases} \ddot x_{\delta} = \ddot x_{0} + \ddot x_{d} \\ \dot x_{\delta} = \dot x_{0} + \dot x_{d} \\ \frac{1}{x_{\delta}} = 1 - x_d \\ \end{cases} \Rightarrow \ddot x_{0} + \ddot x_{d} + \dot x_{0} + \dot x_{d} + 1 - x_d = 1 \Rightarrow \ddot x_{d} + \dot x_{d} - x_d = 0 ⎩ ⎨ ⎧x¨δ=x¨0+x¨dx˙δ=x˙0+x˙dxδ1=1−xd⇒x¨0+x¨d+x˙0+x˙d+1−xd=1⇒x¨d+x˙d−xd=0
二维系统,举个例子
2维空间中,在平衡点附近
{ x ˙ 1 = f 1 ( x 1 , x 2 ) x ˙ 2 = f 2 ( x 1 , x 2 ) ⇒ [ x ˙ 1 d x ˙ 2 d ] = [ ∂ f 1 x 1 ∂ f 1 x 2 ∂ f 2 x 1 ∂ f 2 x 2 ] x = x 0 [ x 1 d x 2 d ] \begin{cases} \dot x_1 = f_1 (x_1,x_2) \\ \dot x_2 = f_2 (x_1,x_2) \\ \end{cases} \Rightarrow \begin{bmatrix} \dot x_{1d} \\ \dot x_{2d} \end{bmatrix} = \begin{bmatrix} \frac{\partial f_1}{x_1} & \frac{\partial f_1}{x_2} \\ \frac{\partial f_2}{x_1} & \frac{\partial f_2}{x_2} \\ \end{bmatrix} _ {x = x_0} \begin{bmatrix} x_{1d} \\ x_{2d} \end{bmatrix} {x˙1=f1(x1,x2)x˙2=f2(x1,x2)⇒[x˙1dx˙2d]=[x1∂f1x1∂f2x2∂f1x2∂f2]x=x0[x1dx2d]
x ¨ + x ˙ + 1 x = 1 \ddot x + \dot x + \frac{1}{x} = 1 x¨+x˙+x1=1
let x 1 = x , x 2 = x ˙ x_1 = x,x_2 = \dot x x1=x,x2=x˙ ,that has
{ x ˙ 1 = x 2 x ˙ 2 = x ¨ = 1 − x ˙ − 1 x = 1 − x 2 − 1 x 1 \begin{cases} \dot x_1 = x_2 \\ \dot x_2 = \ddot x = 1- \dot x - \frac{1}{x} = 1- x_2 - \frac{1}{x_1} \end{cases} {x˙1=x2x˙2=x¨=1−x˙−x1=1−x2−x11
寻找平衡点,令 x ˙ 1 = 0 , x ˙ 2 = 0 \dot x_1 = 0,\dot x_2 = 0 x˙1=0,x˙2=0 ,则有平衡点 x 10 = 1 , x 20 = 0 x_{10} = 1,x_{20} = 0 x10=1,x20=0,
[
x
˙
1
d
x
˙
2
d
]
=
[
0
1
−
(
−
1
x
1
2
)
−
1
]
x
0
[
x
1
d
x
2
d
]
=
[
0
1
1
−
1
]
[
x
1
d
x
2
d
]
\begin{bmatrix} \dot x_{1d} \\ \dot x_{2d} \\ \end{bmatrix} = \begin{bmatrix} 0 & 1 \\ -(-\frac{1}{x_1^2}) & -1 \\ \end{bmatrix} _ {x_0} \begin{bmatrix} x_{1d} \\ x_{2d} \\ \end{bmatrix} = \begin{bmatrix} 0 & 1 \\ 1 & -1 \\ \end{bmatrix} \begin{bmatrix} x_{1d} \\ x_{2d} \end{bmatrix}
[x˙1dx˙2d]=[0−(−x121)1−1]x0[x1dx2d]=[011−1][x1dx2d]
即
{
x
˙
1
d
=
x
2
d
x
˙
2
d
=
x
1
d
−
x
2
d
\begin{cases} \dot x_{1d} = x_{2d} \\ \dot x_{2d} = x_{1d} - x_{2d} \end{cases}
{x˙1d=x2dx˙2d=x1d−x2d
其实我们只需要下半部分
x
˙
2
d
=
x
1
d
−
x
2
d
⇒
x
¨
d
=
x
d
−
x
˙
d
⇒
x
¨
d
+
x
˙
d
−
x
d
=
0
\dot x_{2d} = x_{1d} - x_{2d} \Rightarrow \ddot x_d = x_d - \dot x_d \Rightarrow \ddot x_d + \dot x_d - x_d = 0
x˙2d=x1d−x2d⇒x¨d=xd−x˙d⇒x¨d+x˙d−xd=0
和上面的一维系统的等式是一样的。
总结
线性化公式,
x
−
x
0
→
0
x-x_0 \to 0
x−x0→0
f
(
x
)
=
f
(
x
0
)
+
f
′
(
x
0
)
(
x
−
x
0
)
f(x) = f(x_0) + f'(x_0)(x-x_0)
f(x)=f(x0)+f′(x0)(x−x0)