-
Notifications
You must be signed in to change notification settings - Fork 1
/
example.tex
executable file
·47 lines (39 loc) · 3 KB
/
example.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
\documentclass[12pt]{homework}
\title{Matrix Analysis \quad Homework 6}
\author{Tao}
\email{[email protected]}
\institute{School of Mathematics}
\begin{document}
\maketitle
\section*{1.2.P15}
Suppose that $\mathbf{A}(t) \in M_n$ is a given continuous matrix-valued function and each of the vector valued functions $x_1(t) , \cdots, x_n(t) \in \mathbb{C}^n$ satisfies the system of ordinary differential equations
\begin{equation*}
x^\prime _j (t) = \mathbf{A}(t) x _j (t).
\end{equation*}
Let $\mathbf{X}(t) = [x_1(t) \; \cdots \; x_n(t)]$ and let $W(t) = \det \mathbf{X}(t)$. Use (0.8.10) and (0.8.2.11) and provide details for the following argument:
\begin{equation*}
\begin{aligned}
W^{\prime}(t) &=\sum_{j=1}^{n} \operatorname{det}\left(\mathbf{X}(t) \xleftarrow[j]{} x_{j}^{\prime}(t)\right)=\operatorname{tr}\left[\operatorname{det}\left(\mathbf{X}(t) \xleftarrow[i]{} x_{j}^{\prime}(t)\right)\right]_{i, j=1}^{n} \\
&=\operatorname{tr}\left((\operatorname{adj} \mathbf{X}(t)) \mathbf{X}^{\prime}(t)\right)=\operatorname{tr}((\operatorname{adj} \mathbf{X}(t)) \mathbf{A}(t) \mathbf{X}(t))=W(t) \operatorname{tr} \mathbf{A}(t).
\end{aligned}
\end{equation*}
Thus, $W(t)$ satisfies the scalar differential equation $W^\prime (t) = \operatorname{tr} \mathbf{A}(t)W(t)$, whose solution is \textit{Abel’s formula} for the \textit{Wronskian}
\begin{equation*}
W(t)=W\left(t_{0}\right) e^{\int_{t_{0}}^{t} \operatorname{tr} \mathbf{A}(s) \mathrm{d} s}.
\end{equation*}
Conclude that if the vectors $x_1(t) , \cdots, x_n(t)$ are linearly independent for $t=t_0$, then they are linearly independent for all $t$. How did you use the identity $\operatorname{tr} (\mathbf{B}\mathbf{C}) = \operatorname{tr} (\mathbf{C}\mathbf{B})$ (1.2.P2)?
\begin{solution}
\begin{align*}
W^\prime (t) &= \frac{d}{dt} \det \mathbf{X}(t) \\
\overset{\text{(0.8.10)}}&{=} \sum_{j=1}^{n} \operatorname{det}\left(\mathbf{X}(t) \xleftarrow[j]{} x_{j}^{\prime}(t)\right) \\
&= \operatorname{tr}\left[\operatorname{det}\left(\mathbf{X}(t) \xleftarrow[i]{} x_{j}^{\prime}(t)\right)\right]_{i, j=1}^{n} \\
\overset{\text{(0.8.2.11)}}&{=} \operatorname{tr}\left((\operatorname{adj} \mathbf{X}(t)) \mathbf{X}^{\prime}(t)\right) \\
\overset{x^\prime _j (t) = \mathbf{A}(t) x _j (t)}&{=} \operatorname{tr}((\operatorname{adj} \mathbf{X}(t)) \mathbf{A}(t) \mathbf{X}(t)) \\
\overset{\operatorname{tr} (\mathbf{B}\mathbf{C}) = \operatorname{tr} (\mathbf{C}\mathbf{B})}&{=} \operatorname{tr}( \mathbf{X}(t) (\operatorname{adj} \mathbf{X}(t)) \mathbf{A}(t))\\
&= \det \mathbf{X}(t) \operatorname{tr} \mathbf{A}(t) = W(t) \operatorname{tr} \mathbf{A}(t).
\end{align*}
If the vectors $x_1(t) , \cdots, x_n(t)$ are linearly independent for $t=t_0$, $W(t) = \det \mathbf{X}(t) \neq 0$.
Hence $W(t)=W\left(t_{0}\right) e^{\int_{t_{0}}^{t} \operatorname{tr} \mathbf{A}(s) \mathrm{d} s} \neq 0$.
Therefore $x_1(t) , \cdots, x_n(t)$ are linearly independent for all $t$.
\end{solution}
\end{document}