110 年 國立臺灣聯合大學系統(清華、政治、陽明交通、中央四校聯招)研究所電機類《工程數學(C)》

📄 試題原卷 免費註冊後即可對照原始考卷 PDF免費註冊

第 1 題

Let A=[a1a2a3a4a5]A = \begin{bmatrix} \mathbf{a}_1 & \mathbf{a}_2 & \mathbf{a}_3 & \mathbf{a}_4 & \mathbf{a}_5 \end{bmatrix} be a matrix, and its reduced row echelon form is
R=[10020001100000100000]R = \begin{bmatrix} 1 & 0 & 0 & 2 & 0 \\ 0 & 0 & 1 & 1 & 0 \\ 0 & 0 & 0 & 0 & 1 \\ 0 & 0 & 0 & 0 & 0 \end{bmatrix}.
Which of the following statements is/are true?

(A) a2\mathbf{a}_2 is the 4×14 \times 1 zero vector.
(B) The three column vectors a1,a4\mathbf{a}_1, \mathbf{a}_4 and a5\mathbf{a}_5 are linearly independent.
(C) The column space of RR is the same as the column space of AA.
(D) Let B=[a1a3a2a4]B = \begin{bmatrix} \mathbf{a}_1 & \mathbf{a}_3 \\ \mathbf{a}_2 & \mathbf{a}_4 \end{bmatrix} be a submatrix of AA. The reduced row echelon form of BB is not always a submatrix of RR.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

參考書等級:大學二年級線性代數教材

已知 A=[a1  a2  a3  a4  a5]A=[\mathbf a_1\;\mathbf a_2\;\mathbf a_3\;\mathbf a_4\;\mathbf a_5],其 rref 為

R=[10020001100000100000].R=\begin{bmatrix} 1&0&0&2&0\\ 0&0&1&1&0\\ 0&0&0&0&1\\ 0&0&0&0&0 \end{bmatrix}.

(A) a2\mathbf a_2 為 4×14\times1 零向量

在 rref 中第 2 列全為 0,表示原矩陣第 2 欄不是主軸欄(pivot column),但不代表其全為零向量。若 a2\mathbf a_2 為零向量,其在 rref 中亦會出現全 0 的列;此題的第 2 列已是全 0,故 a2\mathbf a_2 可能是零向量,也可能是其他線性組合。結論:不可必然成立,故 (A) 為 錯。

(B) a1,a4,a5\mathbf a_1,\mathbf a_4,\mathbf a_5 線性獨立

主軸欄位為第 1、3、5 欄,對應 a1,a3,a5\mathbf a_1,\mathbf a_3,\mathbf a_5 為基底。第 4 欄在 rref 中為自由變數列 [2  1  0]T[2\;1\;0]^T,可寫成 a4=2a1+a3\mathbf a_4=2\mathbf a_1+\mathbf a_3. 因此 a4\mathbf a_4 是 a1,a3\mathbf a_1,\mathbf a_3 的線性組合,無法與 a1,a5\mathbf a_1,\mathbf a_5 共同構成獨立集合。結論:錯。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 2 題

Continue from Question 1. Which of the following statements is/are true?

(A) The nonzero rows of RR form a basis for col(AT)col(A^T).
(B) Let CC be the reduced row echelon form of RTR^T. Then CTCC^T C is the identity matrix.
(C) Let EE be a 3×53 \times 5 matrix and
D=[101100111000001]D = \begin{bmatrix} 1 & 0 & 1 & 1 & 0 \\ 0 & 1 & 1 & 1 & 0 \\ 0 & 0 & 0 & 0 & 1 \end{bmatrix}
be the reduced row echelon form of EE. We can obtain the intersection of the right null space of AA and the right null space of EE from RR and DD without knowing AA and EE.
(D) Let QQ be a 5×55 \times 5 matrix and rank(Q)=3rank(Q) = 3. Then rank(AQ)=3rank(AQ) = 3.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題主要考查:

  1. 列運算與列空間、零空間的關係。
  2. 列階梯形矩陣的非零列與基底。
  3. 矩陣乘積的秩不等式。
  4. 右零空間的交集如何轉化為聯立方程組。

依題組脈絡,RR 是由矩陣 AA 經初等列運算得到的列階梯形矩陣,因此

row⁡(R)=row⁡(A).\operatorname{row}(R)=\operatorname{row}(A).

又因為

col⁡(AT)=row⁡(A),\operatorname{col}(A^T)=\operatorname{row}(A),

所以可利用 RR 研究 AA 的列空間與右零空間。


解題方法

列運算不會改變矩陣的列空間,也不會改變齊次方程組的解集合。因此:

Ax=0  ⟺  Rx=0.Ax=0 \iff Rx=0.

若另一矩陣 EE 的簡化列階梯形為 DD,則同理有

Ex=0  ⟺  Dx=0.Ex=0 \iff Dx=0.

故

N(A)∩N(E)={x:Rx=0, Dx=0},N(A)\cap N(E) = \{x:Rx=0,\ Dx=0\},

也就是解聯立矩陣方程

[RD]x=0.\begin{bmatrix} R\\ D \end{bmatrix}x=0.

選項分析

(A)正確

題目聲稱:RR 的非零列構成 col⁡(AT)\operatorname{col}(A^T) 的一組基底。

由列運算的性質,

row⁡(R)=row⁡(A).\operatorname{row}(R)=\operatorname{row}(A).

而矩陣的列空間與其轉置矩陣的行空間相同,因此

col⁡(AT)=row⁡(A).\operatorname{col}(A^T)=\operatorname{row}(A).

所以

col⁡(AT)=row⁡(R).\operatorname{col}(A^T)=\operatorname{row}(R).

列階梯形矩陣的非零列彼此線性獨立,且張成其列空間,因此 RR 的非零列構成 col⁡(AT)\operatorname{col}(A^T) 的一組基底。

故(A)正確。


(B)錯誤

令 CC 為 RTR^T 的簡化列階梯形矩陣。簡化列階梯形的條件只保證:

  • 主元為 11;
  • 主元所在列與行的其他元素為 00;
  • 非零列的主元位置呈階梯狀排列。

這些條件不代表 CC 的列向量彼此正交,也不代表每個列向量的長度為 11。

而

CTC=IC^TC=I

表示 CC 的列向量構成一組標準正交組。這是正交矩陣或 Gram–Schmidt 正交化所涉及的性質,並非簡化列階梯形矩陣必然具有的性質。

例如

C=[1101]C= \begin{bmatrix} 1&1\\ 0&1 \end{bmatrix}

已是簡化列階梯形矩陣以外的典型列階梯形式,但其兩個列向量並不正交,且

CTC=[1112]≠I.C^TC = \begin{bmatrix} 1&1\\ 1&2 \end{bmatrix} \neq I.

因此不能由「CC 是 RTR^T 的簡化列階梯形」推出 CTC=IC^TC=I。

故(B)錯誤。


(C)正確

因為 RR 是 AA 的列等價矩陣,所以

Ax=0  ⟺  Rx=0.Ax=0\iff Rx=0.

又因為 DD 是 EE 的簡化列階梯形矩陣,所以

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 3 題

Let UU and VV be subspaces of Rn\mathbb{R}^n and TT be a linear transformation from Rn\mathbb{R}^n to Rm\mathbb{R}^m. Which of the following sets is/are subspace(s) of Rn\mathbb{R}^n?

(A) {a+b∈Rn:a∈U and b∈V}\{ \mathbf{a} + \mathbf{b} \in \mathbb{R}^n : \mathbf{a} \in U \text{ and } \mathbf{b} \in V \}.
(B) {a∈Rn:a∈U or a∈V}\{ \mathbf{a} \in \mathbb{R}^n : \mathbf{a} \in U \text{ or } \mathbf{a} \in V \}.
(C) {a∈Rn:a∈U and a∈V}\{ \mathbf{a} \in \mathbb{R}^n : \mathbf{a} \in U \text{ and } \mathbf{a} \in V \}.
(D) {a∈Rn:T(a)=0}\{ \mathbf{a} \in \mathbb{R}^n : T(\mathbf{a}) = \mathbf{0} \}.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

判斷集合是否為 Rn\mathbb{R}^n 的子空間,需檢查:

  1. 包含零向量;
  2. 對向量加法封閉;
  3. 對純量乘法封閉。

本題涉及三個重要結構:

  • 子空間的和:U+V={a+b:a∈U,b∈V}U+V=\{\mathbf{a}+\mathbf{b}:\mathbf{a}\in U,\mathbf{b}\in V\};
  • 子空間的交集:U∩V={a:a∈U,a∈V}U\cap V=\{\mathbf{a}:\mathbf{a}\in U,\mathbf{a}\in V\};
  • 線性轉換的核:ker⁡(T)={a:T(a)=0}\ker(T)=\{\mathbf{a}:T(\mathbf{a})=\mathbf{0}\}。

已知 U,VU,V 是子空間,且 TT 是線性轉換。

解題方法

逐一辨認各選項所代表的集合,再利用子空間的基本定理判斷。

選項 (A)

集合為

{a+b∈Rn:a∈U,b∈V}=U+V.\{\mathbf{a}+\mathbf{b}\in\mathbb{R}^n:\mathbf{a}\in U,\mathbf{b}\in V\}=U+V.

取任意 x,y∈U+V\mathbf{x},\mathbf{y}\in U+V,則存在 a1,a2∈U\mathbf{a}_1,\mathbf{a}_2\in U 及 b1,b2∈V\mathbf{b}_1,\mathbf{b}_2\in V,使得

x=a1+b1,y=a2+b2.\mathbf{x}=\mathbf{a}_1+\mathbf{b}_1,\qquad \mathbf{y}=\mathbf{a}_2+\mathbf{b}_2.

由於 U,VU,V 都是子空間,

a1+a2∈U,b1+b2∈V.\mathbf{a}_1+\mathbf{a}_2\in U,\qquad \mathbf{b}_1+\mathbf{b}_2\in V.

因此

x+y=(a1+a2)+(b1+b2)∈U+V.\mathbf{x}+\mathbf{y} =(\mathbf{a}_1+\mathbf{a}_2)+(\mathbf{b}_1+\mathbf{b}_2)\in U+V.

對任意純量 cc,

cx=c(a1+b1)=ca1+cb1.c\mathbf{x}=c(\mathbf{a}_1+\mathbf{b}_1) =c\mathbf{a}_1+c\mathbf{b}_1.

因 ca1∈Uc\mathbf{a}_1\in U 且 cb1∈Vc\mathbf{b}_1\in V,所以 cx∈U+Vc\mathbf{x}\in U+V。

此外,因 0∈U\mathbf{0}\in U 且 0∈V\mathbf{0}\in V,

0=0+0∈U+V.\mathbf{0}=\mathbf{0}+\mathbf{0}\in U+V.

故 U+VU+V 為子空間,選項 (A) 正確。

選項 (B)

集合為

{a∈Rn:a∈U 或 a∈V}=U∪V.\{\mathbf{a}\in\mathbb{R}^n:\mathbf{a}\in U\text{ 或 }\mathbf{a}\in V\} =U\cup V.

兩個子空間的聯集不一定是子空間。以 R2\mathbb{R}^2 為例,令

U=span⁡{(1,0)},V=span⁡{(0,1)}.U=\operatorname{span}\{(1,0)\},\qquad V=\operatorname{span}\{(0,1)\}.

則

(1,0)∈U∪V,(0,1)∈U∪V,(1,0)\in U\cup V,\qquad (0,1)\in U\cup V,

但

(1,0)+(0,1)=(1,1)∉U∪V.(1,0)+(0,1)=(1,1)\notin U\cup V.

因此 U∪VU\cup V 不具加法封閉性,通常不是子空間。選項 (B) 錯誤。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 4 題

Let A,B,A, B, and CC be n×nn \times n matrices for some positive integer nn. Which of the following statements is/are true?

(A) If det⁡(A)=0\det(A) = 0, then det⁡(R)=0\det(R) = 0, where RR is the reduced row echelon form of AA.
(B) If AB=CA=InAB = CA = I_n, then B=CB = C.
(C) If det⁡(AB)≠0\det(AB) \neq 0, then AA is invertible.
(D) rank(AB)=rank(BA)\text{rank}(AB) = \text{rank}(BA).
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查下列矩陣基本性質:

  1. 初等列運算不會改變矩陣的秩,也不會改變矩陣是否可逆。
  2. 方陣若存在左逆矩陣或右逆矩陣,則該矩陣可逆,且左右逆矩陣相同。
  3. det⁡(AB)=det⁡(A)det⁡(B)\det(AB)=\det(A)\det(B)。
  4. 一般而言,矩陣乘積不具交換律,且 rank⁡(AB)\operatorname{rank}(AB) 不一定等於 rank⁡(BA)\operatorname{rank}(BA)。

選項分析

(A) 正確

若 det⁡(A)=0\det(A)=0,則 AA 不可逆,亦即

rank⁡(A)<n.\operatorname{rank}(A)<n.

將 AA 化為 reduced row echelon form RR 只使用初等列運算,而初等列運算不改變矩陣的秩,因此

rank⁡(R)=rank⁡(A)<n.\operatorname{rank}(R)=\operatorname{rank}(A)<n.

所以 RR 不是滿秩方陣,必有

det⁡(R)=0.\det(R)=0.

因此選項 (A) 正確。


(B) 正確

已知

AB=In,CA=In.AB=I_n,\qquad CA=I_n.

利用矩陣乘法結合律:

C=CIn=C(AB)=(CA)B=InB=B.C=C I_n=C(AB)=(CA)B=I_nB=B.

因此

B=C.B=C.

故選項 (B) 正確。

這也說明方陣的左逆矩陣與右逆矩陣若同時存在,必定相同。


(C) 正確

由行列式乘法公式,

det⁡(AB)=det⁡(A)det⁡(B).\det(AB)=\det(A)\det(B).

題目給定

det⁡(AB)≠0.\det(AB)\neq 0.

因此 det⁡(A)\det(A) 與 det⁡(B)\det(B) 都不可能為 00,特別有

det⁡(A)≠0.\det(A)\neq 0.

方陣行列式非零等價於矩陣可逆,所以 AA 可逆。

故選項 (C) 正確。


🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 5 題

Let VV be the vector space of all A∈Rm×rA \in \mathbb{R}^{m \times r}, with the operations of matrix addition and multiplication of a matrix by a real scalar. Let TT be a linear transformation from VV to VV, and {B1,B2,…,Bn}\{B_1, B_2, \dots, B_n\} be a basis for VV. Suppose UU is a linear transformation from VV to Rn\mathbb{R}^n given by U(c1B1+c2B2+⋯+cnBn)=[c1 c2 … cn]TU(c_1B_1 + c_2B_2 + \dots + c_nB_n) = [c_1 \ c_2 \ \dots \ c_n]^T. Which of the following statements is/are true?

(A) Let {A1,A2,…,Ak}\{A_1, A_2, \dots, A_k\} be linearly independent over R\mathbb{R}. The set {U(A1),U(A2),…,U(Ak)}\{U(A_1), U(A_2), \dots, U(A_k)\} can be linearly dependent over R\mathbb{R}.
(B) Suppose the dimension of the range space of TT is kk over R\mathbb{R}. Then {T(B1),T(B2),…,T(Bk)}\{T(B_1), T(B_2), \dots, T(B_k)\} are linearly independent over R\mathbb{R}.
(C) n=mn = m.
(D) Let c=U(B1)c = U(B_1) and a=U(T(B1))a = U(T(B_1)), then aaT=ccTa a^T = c c^T.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

解題關鍵

  • V=Rm×rV=\mathbb R^{m\times r},dim⁡V=mr\dim V=m r,基底 {B1,…,Bn}\{B_1,\dots ,B_n\},其中 n=mrn=m r。
  • U:V→RnU:V\to\mathbb R^n 為座標映射,對任意 A=∑i=1nciBiA=\sum_{i=1}^n c_iB_i,U(A)=[c1  c2  …  cn]TU(A)=[c_1\;c_2\;\dots\;c_n]^T。
  • UU 為線性雙射(同構),保留線性獨立性與相依性。

(A)

若 {A1,…,Ak}\{A_1,\dots ,A_k\} 線性無關,則對任意係數 cic_i,∑ciAi=0  ⟺  ci=0\sum c_iA_i=0\iff c_i=0。
施以 UU 得 ∑ciU(Ai)=U(0)=0\sum c_iU(A_i)=U(0)=0,同樣只有平凡解,所以 {U(Ai)}\{U(A_i)\} 亦線性無關。
結論:錯(不可能相依)。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 6 題

For the matrix
A=[10−1][1−10]+[111][111]+[1−20][00−1]A = \begin{bmatrix} 1 \\ 0 \\ -1 \end{bmatrix} \begin{bmatrix} 1 & -1 & 0 \end{bmatrix} + \begin{bmatrix} 1 \\ 1 \\ 1 \end{bmatrix} \begin{bmatrix} 1 & 1 & 1 \end{bmatrix} + \begin{bmatrix} 1 \\ -2 \\ 0 \end{bmatrix} \begin{bmatrix} 0 & 0 & -1 \end{bmatrix}
which of the following statements is/are true?

(A) rank(A)=3\text{rank}(A) = 3.
(B) The sum of eigenvalues of AA is 6.
(C) AA is similar to B=[1−1402−1003]B = \begin{bmatrix} 1 & -1 & 4 \\ 0 & 2 & -1 \\ 0 & 0 & 3 \end{bmatrix}.
(D) The system of linear equations Ax=[−111]TA \mathbf{x} = \begin{bmatrix} -1 \\ 1 \\ 1 \end{bmatrix}^T has a solution.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查:

  • 外積矩陣的計算;
  • 矩陣的秩與行列式;
  • 特徵值總和與跡的關係:
    λ1+λ2+λ3=tr⁡(A);\lambda_1+\lambda_2+\lambda_3=\operatorname{tr}(A);
  • 相似矩陣具有相同的特徵值、行列式與跡;
  • 可逆矩陣所對應的線性方程組必有唯一解。

解題方法

先將三個外積矩陣分別計算,再相加。

[10−1][1−10]=[1−10000−110],\begin{bmatrix}1\\0\\-1\end{bmatrix} \begin{bmatrix}1&-1&0\end{bmatrix} = \begin{bmatrix} 1&-1&0\\ 0&0&0\\ -1&1&0 \end{bmatrix}, [111][111]=[111111111],\begin{bmatrix}1\\1\\1\end{bmatrix} \begin{bmatrix}1&1&1\end{bmatrix} = \begin{bmatrix} 1&1&1\\ 1&1&1\\ 1&1&1 \end{bmatrix}, [1−20][00−1]=[00−1002000].\begin{bmatrix}1\\-2\\0\end{bmatrix} \begin{bmatrix}0&0&-1\end{bmatrix} = \begin{bmatrix} 0&0&-1\\ 0&0&2\\ 0&0&0 \end{bmatrix}.

因此

A=[200113021].A= \begin{bmatrix} 2&0&0\\ 1&1&3\\ 0&2&1 \end{bmatrix}.

選項分析

(A) rank⁡(A)=3\operatorname{rank}(A)=3

計算行列式:

det⁡(A)=2∣1321∣=2(1−6)=−10≠0.\det(A) = 2\begin{vmatrix} 1&3\\ 2&1 \end{vmatrix} = 2(1-6) =-10\neq 0.

因此 AA 可逆,為滿秩矩陣:

rank⁡(A)=3.\operatorname{rank}(A)=3.

故 (A) 正確。


(B) AA 的特徵值總和為 66

矩陣特徵值總和等於矩陣的跡:

λ1+λ2+λ3=tr⁡(A).\lambda_1+\lambda_2+\lambda_3 =\operatorname{tr}(A).

由 AA 的對角線元素得

tr⁡(A)=2+1+1=4.\operatorname{tr}(A)=2+1+1=4.

因此特徵值總和為 44,並非 66。

故 (B) 錯誤。


🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 7 題

For a non-zero matrix A∈Rm×nA \in \mathbb{R}^{m \times n}, define σ=max⁡∥x∥=1∥Ax∥\sigma = \max_{\| \mathbf{x} \|=1} \| A \mathbf{x} \| (1)
the maximum Frobenius norm ∥Ax∥\|A \mathbf{x}\| over all unit-norm vectors x\mathbf{x}. Which of the following statements is/are true?

(A) σ2\sigma^2 is the maximal eigenvalue of AATA A^T.
(B) If x0\mathbf{x}_0 is an optimal solution to equation (1), i.e., ∥Ax0∥=σ\| A \mathbf{x}_0 \| = \sigma, then there exists a vector x1\mathbf{x}_1 such that Ax1=0A \mathbf{x}_1 = \mathbf{0} and x1Tx0≠0\mathbf{x}_1^T \mathbf{x}_0 \neq 0.
(C) Assume m=nm = n and AA is nonsingular. Then 1σ=max⁡∥x∥=1∥A−1x∥\frac{1}{\sigma} = \max_{\| \mathbf{x} \|=1} \| A^{-1} \mathbf{x} \|.
(D) Assume m=nm = n. Then In+AI_n + A is singular only if σ≥1\sigma \ge 1.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

解題要點

  • σ=max⁡∥x∥=1∥Ax∥\sigma=\max_{\|x\|=1}\|Ax\| 為 AA 的算子範數(即最大奇異值),滿足
    σ2=λmax⁡(ATA)=λmax⁡(AAT).\sigma^{2}=\lambda_{\max}(A^{T}A)=\lambda_{\max}(AA^{T}).

  • 右奇異向量 viv_{i}(∥vi∥=1\|v_{i}\|=1)構成 Rn\mathbb{R}^{n} 的正交基,且
    Avi=σiui,σi≥0Av_{i}=\sigma_{i}u_{i},\qquad\sigma_{i}\ge0
    其中 σi\sigma_{i} 為奇異值,uiu_{i} 為左奇異向量。
    若 σi=0\sigma_{i}=0,則 vi∈ker⁡Av_{i}\in\ker A,且必與所有 σj>0\sigma_{j}>0 對應的 vjv_{j} 正交。


(A)

σ2=λmax⁡(AAT)\sigma^{2}=\lambda_{\max}(AA^{T})
由上式直接得證。成立。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 8 題

Let VV together with the inner product ⟨⋅,⋅⟩V\langle \cdot, \cdot \rangle_V be an inner product space over R\mathbb{R}. For a collection of NN linearly independent vectors {w1,…,wN}\{\mathbf{w}_1, \dots, \mathbf{w}_N\} in VV, let M=[mi,j]∈RN×NM = [m_{i,j}] \in \mathbb{R}^{N \times N} be the corresponding Gram matrix, i.e., the (i,j)(i, j)-th entry of MM is given by mi,j=⟨wi,wj⟩Vm_{i,j} = \langle \mathbf{w}_i, \mathbf{w}_j \rangle_V. Define the following set of real-valued functions F:={f:V→R}F := \{ f: V \to \mathbb{R} \}
f(x)=∑n=1Nαn⟨wn,x⟩Vf(\mathbf{x}) = \sum_{n=1}^N \alpha_n \langle \mathbf{w}_n, \mathbf{x} \rangle_V, where α1,…,αN∈R\alpha_1, \dots, \alpha_N \in \mathbb{R}.
FF is a vector space over R\mathbb{R} when combined with standard addition and scalar multiplication of real-valued functions. For any f(x)=∑n=1Nαn⟨wn,x⟩Vf(\mathbf{x}) = \sum_{n=1}^N \alpha_n \langle \mathbf{w}_n, \mathbf{x} \rangle_V and g(y)=∑m=1Nβm⟨wm,y⟩Vg(\mathbf{y}) = \sum_{m=1}^N \beta_m \langle \mathbf{w}_m, \mathbf{y} \rangle_V in the vector space FF, define
(f,g)F=∑n=1N∑m=1Nαnβm⟨wn,wm⟩V(f, g)_F = \sum_{n=1}^N \sum_{m=1}^N \alpha_n \beta_m \langle \mathbf{w}_n, \mathbf{w}_m \rangle_V.
Which of the following statements is/are true?

(A) MM is positive definite.
(B) If m1,1=2m_{1,1} = 2, m2,2=1m_{2,2} = 1, and m1,2=−12m_{1,2} = -\frac{1}{2}, then ∥w1−w2∥V=4\| \mathbf{w}_1 - \mathbf{w}_2 \|_V = 4, where ∥⋅∥V\| \cdot \|_V is the vector norm induced by the inner product ⟨⋅,⋅⟩V\langle \cdot, \cdot \rangle_V.
(C) (f,g)F(f, g)_F is an inner product for elements f,g∈Ff, g \in F.
(D) Given g(y)=⟨w1,y⟩Vg(\mathbf{y}) = \langle \mathbf{w}_1, \mathbf{y} \rangle_V, we have (f,g)F=f(w1)(f, g)_F = f(\mathbf{w}_1) for all f∈Ff \in F.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查三個重點:

  1. 線性獨立向量的 Gram 矩陣是否為正定矩陣。
  2. 內積誘導的向量範數計算。
  3. 以函數係數與 Gram 矩陣定義的 (⋅,⋅)F(\cdot,\cdot)_F 是否構成內積,以及其與函數值的關係。

Gram 矩陣為

M=[mi,j],mi,j=⟨wi,wj⟩V.M=[m_{i,j}],\qquad m_{i,j}=\langle \mathbf w_i,\mathbf w_j\rangle_V.

對任意 c=(c1,…,cN)T∈RN\mathbf c=(c_1,\dots,c_N)^T\in\mathbb R^N,

cTMc=∑i=1N∑j=1Ncicj⟨wi,wj⟩V=∥∑i=1Nciwi∥V2.\mathbf c^T M\mathbf c = \sum_{i=1}^N\sum_{j=1}^N c_i c_j \langle \mathbf w_i,\mathbf w_j\rangle_V = \left\|\sum_{i=1}^N c_i\mathbf w_i\right\|_V^2.

這個公式是判斷 Gram 矩陣正定性的關鍵。


解題方法

逐一檢查各選項:

  • 先利用向量線性獨立性判斷 Gram 矩陣。
  • 利用
∥u−v∥V2=⟨u,u⟩V+⟨v,v⟩V−2⟨u,v⟩V\|\mathbf u-\mathbf v\|_V^2 = \langle \mathbf u,\mathbf u\rangle_V +\langle \mathbf v,\mathbf v\rangle_V -2\langle \mathbf u,\mathbf v\rangle_V

計算選項 B。

  • 將 (f,g)F(f,g)_F 改寫成兩個向量線性組合的內積,驗證內積公理。
  • 代入 g(y)=⟨w1,y⟩Vg(\mathbf y)=\langle\mathbf w_1,\mathbf y\rangle_V 的係數,檢查選項 D。

選項分析

(A) MM 是正定矩陣

對任意非零向量 c∈RN\mathbf c\in\mathbb R^N,

cTMc=∥∑i=1Nciwi∥V2.\mathbf c^T M\mathbf c = \left\|\sum_{i=1}^N c_i\mathbf w_i\right\|_V^2.

由於 {w1,…,wN}\{\mathbf w_1,\dots,\mathbf w_N\} 線性獨立,當 c≠0\mathbf c\neq\mathbf 0 時,

∑i=1Nciwi≠0.\sum_{i=1}^N c_i\mathbf w_i\neq \mathbf 0.

因此

∥∑i=1Nciwi∥V2>0,\left\|\sum_{i=1}^N c_i\mathbf w_i\right\|_V^2>0,

故

cTMc>0(c≠0).\mathbf c^T M\mathbf c>0\qquad(\mathbf c\neq\mathbf 0).

所以 MM 為正定矩陣。

(A) 正確。


(B) ∥w1−w2∥V=4\|\mathbf w_1-\mathbf w_2\|_V=4

由題目條件,

⟨w1,w1⟩V=2,⟨w2,w2⟩V=1,⟨w1,w2⟩V=−12.\langle\mathbf w_1,\mathbf w_1\rangle_V=2,\qquad \langle\mathbf w_2,\mathbf w_2\rangle_V=1,\qquad \langle\mathbf w_1,\mathbf w_2\rangle_V=-\frac12.

因此

∥w1−w2∥V2=⟨w1−w2,w1−w2⟩V=⟨w1,w1⟩V−2⟨w1,w2⟩V+⟨w2,w2⟩V=2−2(−12)+1=4.\begin{aligned} \|\mathbf w_1-\mathbf w_2\|_V^2 &= \langle\mathbf w_1-\mathbf w_2,\mathbf w_1-\mathbf w_2\rangle_V\\ &= \langle\mathbf w_1,\mathbf w_1\rangle_V -2\langle\mathbf w_1,\mathbf w_2\rangle_V +\langle\mathbf w_2,\mathbf w_2\rangle_V\\ &= 2-2\left(-\frac12\right)+1\\ &=4. \end{aligned}

所以

∥w1−w2∥V=2,\|\mathbf w_1-\mathbf w_2\|_V=2,

而不是 44。

(B) 錯誤。


(C) (f,g)F(f,g)_F 是 FF 上的內積

令

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 9 題

Which of the following statements is/are true?

(A) Let x\mathbf{x} and y\mathbf{y} be two non-zero vectors in Rn\mathbb{R}^n. The matrix xyT\mathbf{x} \mathbf{y}^T is diagonalizable if yTx≠0\mathbf{y}^T \mathbf{x} \neq 0.
(B) If A∈Rn×nA \in \mathbb{R}^{n \times n} is symmetric and positive definite, then A+A−1−2InA + A^{-1} - 2I_n is positive semi-definite.
(C) For u,v∈Rn\mathbf{u}, \mathbf{v} \in \mathbb{R}^n such that uTv≠0\mathbf{u}^T \mathbf{v} \neq 0, Rn=(span{u})⊕(span{v})\mathbb{R}^n = (\text{span}\{\mathbf{u}\}) \oplus (\text{span}\{\mathbf{v}\}), where ⊕\oplus is the operation of direct sum of two vector spaces.
(D) For the real matrix Q=In−uuTQ = I_n - \mathbf{u} \mathbf{u}^T, where u∈Rn\mathbf{u} \in \mathbb{R}^n, we have det⁡(Q)+rank(Q)=n\det(Q) + \text{rank}(Q) = n.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題綜合考查:

  • 秩一矩陣 xyT\mathbf{x}\mathbf{y}^T 的特徵值與可對角化性。
  • 對稱正定矩陣的譜分解,以及正半定矩陣的判定。
  • 直和的定義與維度公式。
  • In−uuTI_n-\mathbf{u}\mathbf{u}^T 的特徵值、行列式與秩。
  • 「以上皆非」選項的判斷。

解題方法與選項分析

選項 (A)

令

M=xyT.M=\mathbf{x}\mathbf{y}^T.

由矩陣乘法可得

Mx=xyTx=(yTx)x.M\mathbf{x} =\mathbf{x}\mathbf{y}^T\mathbf{x} =(\mathbf{y}^T\mathbf{x})\mathbf{x}.

因此,x\mathbf{x} 是 MM 的特徵向量,對應特徵值

λ=yTx.\lambda=\mathbf{y}^T\mathbf{x}.

又因為 MM 的秩最多為 11,當 yTx≠0\mathbf{y}^T\mathbf{x}\neq 0 時,MM 的非零特徵值只有 λ\lambda,其餘 n−1n-1 個特徵值為 00。

此外,

M2=xyTxyT=(yTx)M=λM.M^2 =\mathbf{x}\mathbf{y}^T\mathbf{x}\mathbf{y}^T =(\mathbf{y}^T\mathbf{x})M =\lambda M.

所以 MM 滿足多項式

p(t)=t(t−λ).p(t)=t(t-\lambda).

由於 λ≠0\lambda\neq 0,00 與 λ\lambda 是兩個相異根,故 MM 的最小多項式沒有重根,因此 MM 可對角化。

所以選項 (A) 正確。


選項 (B)

因為 AA 為實對稱正定矩陣,所以存在正交矩陣 PP,使得

A=Pdiag⁡(λ1,λ2,…,λn)PT,A=P\operatorname{diag}(\lambda_1,\lambda_2,\ldots,\lambda_n)P^T,

其中每個特徵值皆滿足 λi>0\lambda_i>0。

因此

A−1=Pdiag⁡(1λ1,…,1λn)PT.A^{-1} =P\operatorname{diag}\left(\frac1{\lambda_1},\ldots,\frac1{\lambda_n}\right)P^T.

故

A+A−1−2In=Pdiag⁡(λi+1λi−2)i=1nPT.A+A^{-1}-2I_n = P\operatorname{diag}\left( \lambda_i+\frac1{\lambda_i}-2 \right)_{i=1}^nP^T.

對每個 λi>0\lambda_i>0,

λi+1λi−2=(λi−1)2λi≥0.\lambda_i+\frac1{\lambda_i}-2 = \frac{(\lambda_i-1)^2}{\lambda_i} \geq 0.

所以 A+A−1−2InA+A^{-1}-2I_n 的所有特徵值皆非負,因而為正半定矩陣。

所以選項 (B) 正確。


選項 (C)

由 uTv≠0\mathbf{u}^T\mathbf{v}\neq 0 可知 u\mathbf{u} 與 v\mathbf{v} 都是非零向量,但這個條件不能保證它們線性獨立。

例如取

u=v≠0,\mathbf{u}=\mathbf{v}\neq \mathbf{0},

則

uTv=uTu=∥u∥2≠0,\mathbf{u}^T\mathbf{v} =\mathbf{u}^T\mathbf{u} =\|\mathbf{u}\|^2\neq 0,

但此時

span⁡{u}=span⁡{v},\operatorname{span}\{\mathbf{u}\} = \operatorname{span}\{\mathbf{v}\},
🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 10 題

Let P∈Rn×nP \in \mathbb{R}^{n \times n} be the orthogonal projection matrix, with respect to the Euclidean inner product, of vectors in Rn\mathbb{R}^n onto col(A) for some nonzero matrix A∈Rn×kA \in \mathbb{R}^{n \times k}. Which of the following statements is/are true?

(A) tr(PTP)≥rank(P)\text{tr}(P^T P) \ge \text{rank}(P).
(B) PP can have an eigenvalue larger than 2.
(C) PxxTP \mathbf{x} \mathbf{x}^T for all x∈Rn\mathbf{x} \in \mathbb{R}^n.
(D) If k<nk < n and b∈Rn\mathbf{b} \in \mathbb{R}^n, the system of linear equations Ax=bA \mathbf{x} = \mathbf{b} has a unique least squares solution.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

正交投影矩陣 PP 將任意向量投影到 col⁡(A)\operatorname{col}(A)。因此具有以下性質:

PT=P,P2=P.P^T=P,\qquad P^2=P.

其中 PT=PP^T=P 表示對稱,P2=PP^2=P 表示冪等。若令

r=rank⁡(P)=dim⁡(col⁡(A)),r=\operatorname{rank}(P)=\dim(\operatorname{col}(A)),

則 PP 的特徵值只能是 00 或 11,且特徵值 11 的個數為 rr。


解題方法

利用正交投影矩陣的對稱性與冪等性,逐一檢驗各選項。特別是:

PTP=P2=P.P^TP=P^2=P.

因此

tr⁡(PTP)=tr⁡(P).\operatorname{tr}(P^TP)=\operatorname{tr}(P).

而投影矩陣的跡等於其秩:

tr⁡(P)=rank⁡(P).\operatorname{tr}(P)=\operatorname{rank}(P).

選項分析

(A) tr⁡(PTP)≥rank⁡(P)\operatorname{tr}(P^TP)\ge \operatorname{rank}(P)

由 PT=PP^T=P 且 P2=PP^2=P,

PTP=P2=P.P^TP=P^2=P.

因此

tr⁡(PTP)=tr⁡(P)=rank⁡(P).\operatorname{tr}(P^TP) =\operatorname{tr}(P) =\operatorname{rank}(P).

所以不等式成立,而且實際上是等號成立。

故 (A) 正確。


(B) PP 可以有大於 22 的特徵值

設 λ\lambda 為 PP 的特徵值,且 v≠0\mathbf{v}\ne\mathbf{0} 滿足

Pv=λv.P\mathbf{v}=\lambda\mathbf{v}.

由 P2=PP^2=P,

P2v=Pv.P^2\mathbf{v}=P\mathbf{v}.

所以

λ2v=λv,\lambda^2\mathbf{v}=\lambda\mathbf{v},

即

λ(λ−1)=0.\lambda(\lambda-1)=0.

因此

λ=0或λ=1.\lambda=0\quad\text{或}\quad \lambda=1.

不可能存在大於 22 的特徵值。

故 (B) 錯誤。


(C)

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 11 題

Consider the first order differential equation xy′(x)=y(x)−x2+(y(x))2xy'(x) = y(x) - \sqrt{x^2 + (y(x))^2} with initial condition y(2)=0y(2) = 0. Which of the following statements is/are true?

(A) It is a linear differential equation for the dependent variable yy.
(B) y(x)y(x) is a parabolic function of xx.
(C) y′(x)=−xy'(x) = -x.
(D) y′′(x)=−12y''(x) = -\frac{1}{2}.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查:

  • 一階微分方程的線性判斷。
  • 齊次型微分方程的代換 v=yxv=\dfrac{y}{x}。
  • 雙曲函數積分公式
    ∫dv1+v2=sinh⁡−1v.\int \frac{dv}{\sqrt{1+v^2}}=\sinh^{-1}v.
  • 由解函數判斷 y′(x)y'(x)、y′′(x)y''(x) 與圖形類型。

解題方法

原方程為

xy′(x)=y(x)−x2+y(x)2,y(2)=0.xy'(x)=y(x)-\sqrt{x^2+y(x)^2}, \qquad y(2)=0.

因初始點 x=2>0x=2>0,在此解的考慮範圍內取 x>0x>0。將方程除以 xx:

y′=yx−1+(yx)2.y'= \frac{y}{x}-\sqrt{1+\left(\frac{y}{x}\right)^2}.

右側只含有 yx\dfrac{y}{x},因此使用齊次型代換

v=yx,y=xv.v=\frac{y}{x},\qquad y=xv.

微分得

y′=v+xdvdx.y'=v+x\frac{dv}{dx}.

代回原方程:

x(v+xdvdx)=xv−x1+v2.x\left(v+x\frac{dv}{dx}\right) =xv-x\sqrt{1+v^2}.

消去 xvxv 後得到

x2dvdx=−x1+v2,x^2\frac{dv}{dx}=-x\sqrt{1+v^2},

即

dv1+v2=−dxx.\frac{dv}{\sqrt{1+v^2}}=-\frac{dx}{x}.

兩邊積分:

sinh⁡−1v=−ln⁡x+C.\sinh^{-1}v=-\ln x+C.

由初始條件 y(2)=0y(2)=0,可得

v(2)=y(2)2=0.v(2)=\frac{y(2)}{2}=0.

因此

0=−ln⁡2+C,C=ln⁡2.0=-\ln 2+C, \qquad C=\ln 2.

所以

sinh⁡−1v=ln⁡2x.\sinh^{-1}v=\ln\frac{2}{x}.

取雙曲正弦:

v=sinh⁡(ln⁡2x).v=\sinh\left(\ln\frac{2}{x}\right).

利用

sinh⁡t=et−e−t2,\sinh t=\frac{e^t-e^{-t}}{2},

可得

v=2x−x22=1x−x4.v =\frac{\frac{2}{x}-\frac{x}{2}}{2} =\frac{1}{x}-\frac{x}{4}.

由 y=xvy=xv:

y=x(1x−x4)=1−x24.y=x\left(\frac{1}{x}-\frac{x}{4}\right) =1-\frac{x^2}{4}.

故解函數為

y(x)=1−x24.\boxed{y(x)=1-\frac{x^2}{4}}.

其導數為

y′(x)=−x2,y'(x)=-\frac{x}{2},
🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 12 題

For the following homogeneous second order linear differential equation
(x2−1)y′′(x)−2xy′(x)+2y(x)=0(x^2 - 1)y''(x) - 2xy'(x) + 2y(x) = 0
for x>1x > 1, given one solution y1(x)=xy_1(x) = x, the other linearly independent solution y2(x)y_2(x) can then be derived by setting y2(x)=v(x)y1(x)y_2(x) = v(x)y_1(x). Which of the following statements is/are true?

(A) (x3+x)ν′′(x)+2ν′(x)=0(x^3 + x)\nu''(x) + 2\nu'(x) = 0.
(B) ν(x)\nu(x) with ν′(x)=x2x2−1\nu'(x) = \frac{x^2}{x^2-1} corresponds to one of the possible solutions.
(C) ν(x)\nu(x) with ν′(x)=xx2−1\nu'(x) = \frac{x}{x^2-1} corresponds to one of the possible solutions.
(D) y(x)=1+x+x2y(x) = 1 + x + x^2 is one of the possible solutions.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

對方程式

(x2−1)y′′−2xy′+2y=0,x>1(x^{2}-1)y''-2xy'+2y=0,\qquad x>1

先除以 x2−1x^{2}-1 得標準形

y′′−2xx2−1y′+2x2−1y=0,y''-\frac{2x}{x^{2}-1}y'+\frac{2}{x^{2}-1}y=0,

故

P(x)=−2xx2−1,Q(x)=2x2−1.P(x)=-\frac{2x}{x^{2}-1},\qquad Q(x)=\frac{2}{x^{2}-1}.

已知解 y1=xy_{1}=x。令 y2=v(x)y1=v(x)xy_{2}=v(x)y_{1}=v(x)x。
使用降階公式

v′′y1+(2y1′+Py1)v′=0.v''y_{1}+\bigl(2y_{1}'+P y_{1}\bigr)v'=0 .

帶入 y1=x,  y1′=1y_{1}=x,\;y_{1}'=1:

xv′′+(2−2x2x2−1)v′=0⟹(x3−x)v′′−2v′=0.xv''+\Bigl(2-\frac{2x^{2}}{x^{2}-1}\Bigr)v'=0 \Longrightarrow (x^{3}-x)v''-2v'=0 .

設 w=v′w=v',則

(x3−x)w′−2w=0  ⟹  w′w=2x3−x=2x(x2−1).(x^{3}-x)w'-2w=0\;\Longrightarrow\; \frac{w'}{w}=\frac{2}{x^{3}-x} =\frac{2}{x(x^{2}-1)} .

部分分式分解

2x(x2−1)=−2x+2xx2−1,\frac{2}{x(x^{2}-1)}=-\frac{2}{x}+\frac{2x}{x^{2}-1},

積分得到

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 13 題

Continue from Question 12. Solve the following non-homogeneous second order linear differential equation
(x2−1)y′′(x)−2xy′(x)+2y(x)=x2−1(x^2 - 1)y''(x) - 2xy'(x) + 2y(x) = x^2 - 1
for x>1x > 1 using the variation of parameters, i.e., set the particular solution as
yp(x)=u1(x)y1(x)+u2(x)y2(x)y_p(x) = u_1(x)y_1(x) + u_2(x)y_2(x)
where y1(x)y_1(x) and y2(x)y_2(x) are homogeneous solutions from Question 12. Which of the following statements is/are true?

(A) u1′(x)=−x2+1x2−1u_1'(x) = -\frac{x^2+1}{x^2-1}
(B) u1′(x)=−(x2+1)u_1'(x) = -(x^2+1)
(C) u2′(x)=x2−1x2−1u_2'(x) = \frac{x^2-1}{x^2-1}
(D) u2′(x)=xu_2'(x) = x
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查二階非齊次線性微分方程的參數變易法。標準形式為

y′′+P(x)y′+Q(x)y=g(x),y''+P(x)y'+Q(x)y=g(x),

若齊次方程的兩個線性獨立解為 y1,y2y_1,y_2,則設

yp=u1y1+u2y2,y_p=u_1y_1+u_2y_2,

並附加條件

u1′y1+u2′y2=0.u_1'y_1+u_2'y_2=0.

代入後可得

u1′y1′+u2′y2′=g(x).u_1'y_1'+u_2'y_2'=g(x).

若

W(y1,y2)=∣y1y2y1′y2′∣=y1y2′−y1′y2,W(y_1,y_2)= \begin{vmatrix} y_1&y_2\\ y_1'&y_2' \end{vmatrix} =y_1y_2'-y_1'y_2,

則

u1′=−y2gW,u2′=y1gW.u_1'=-\frac{y_2g}{W},\qquad u_2'=\frac{y_1g}{W}.

解題方法

原方程為

(x2−1)y′′−2xy′+2y=x2−1.(x^2-1)y''-2xy'+2y=x^2-1.

除以 x2−1x^2-1,且因 x>1x>1,故分母不為零:

y′′−2xx2−1y′+2x2−1y=1.y''-\frac{2x}{x^2-1}y' +\frac{2}{x^2-1}y=1.

由第 12 題,取兩個齊次解為

y1=x,y2=x2+1.y_1=x,\qquad y_2=x^2+1.

其導數為

y1′=1,y2′=2x.y_1'=1,\qquad y_2'=2x.

因此 Wronskian 為

W=∣xx2+112x∣=x(2x)−(x2+1)=x2−1.W= \begin{vmatrix} x&x^2+1\\ 1&2x \end{vmatrix} =x(2x)-(x^2+1) =x^2-1.

右端函數為 g(x)=1g(x)=1,所以參數變易法給出

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 14 題

The following second order system
{x′′(t)=−40x(t)+8y(t)y′′(t)=12x(t)−60y(t)\begin{cases} x''(t) = -40x(t) + 8y(t) \\ y''(t) = 12x(t) - 60y(t) \end{cases} (2)
can be transformed into an equivalent first order system
{x1′(t)=x2(t)x2′(t)=−40x1(t)+8x3(t)x3′(t)=x4(t)x4′(t)=12x2(t)−60x3(t)=A{x1(t)x2(t)x3(t)x4(t)\begin{cases} x_1'(t) = x_2(t) \\ x_2'(t) = -40x_1(t) + 8x_3(t) \\ x_3'(t) = x_4(t) \\ x_4'(t) = 12x_2(t) - 60x_3(t) \end{cases} = A \begin{cases} x_1(t) \\ x_2(t) \\ x_3(t) \\ x_4(t) \end{cases}
for some real matrix AA by introducing the four functions x1(t)=x(t),x2(t)=x′(t),x3(t)=y(t)x_1(t) = x(t), x_2(t) = x'(t), x_3(t) = y(t) and x4(t)=y′(t)x_4(t) = y'(t). Which of the following statements is/are true?

(A) A=[0100−400800001120−600]A = \begin{bmatrix} 0 & 1 & 0 & 0 \\ -40 & 0 & 8 & 0 \\ 0 & 0 & 0 & 1 \\ 12 & 0 & -60 & 0 \end{bmatrix}
(B) A=[0100−400800001012−600]A = \begin{bmatrix} 0 & 1 & 0 & 0 \\ -40 & 0 & 8 & 0 \\ 0 & 0 & 0 & 1 \\ 0 & 12 & -60 & 0 \end{bmatrix}
(C) -36 is one of the eigenvalues of AA.
(D) 8 is one of the eigenvalues of AA.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查兩個重點:

  1. 將二階微分方程組改寫成一階線性系統。
  2. 由特徵方程判斷矩陣的 eigenvalues。

若一階系統寫成

x′(t)=Ax(t),\mathbf{x}'(t)=A\mathbf{x}(t),

則矩陣 AA 的 eigenvalue λ\lambda 必須滿足

det⁡(λI−A)=0.\det(\lambda I-A)=0.

解題方法一:建立矩陣 AA

題目定義

x1=x,x2=x′,x3=y,x4=y′.x_1=x,\qquad x_2=x',\qquad x_3=y,\qquad x_4=y'.

因此各分量的導數為

x1′=x2,x_1'=x_2, x2′=x′′=−40x+8y=−40x1+8x3,x_2'=x''=-40x+8y=-40x_1+8x_3, x3′=x4,x_3'=x_4, x4′=y′′=12x−60y=12x1−60x3.x_4'=y''=12x-60y=12x_1-60x_3.

故

[x1′x2′x3′x4′]=[0100−400800001120−600][x1x2x3x4].\begin{bmatrix} x_1'\\ x_2'\\ x_3'\\ x_4' \end{bmatrix} = \begin{bmatrix} 0&1&0&0\\ -40&0&8&0\\ 0&0&0&1\\ 12&0&-60&0 \end{bmatrix} \begin{bmatrix} x_1\\ x_2\\ x_3\\ x_4 \end{bmatrix}.

所以選項 A 的矩陣正確。


解題方法二:求矩陣的 eigenvalues

對原二階方程組設

x(t)=Xeλt,y(t)=Yeλt.x(t)=Xe^{\lambda t},\qquad y(t)=Ye^{\lambda t}.

則

x′′=λ2Xeλt,y′′=λ2Yeλt.x''=\lambda^2Xe^{\lambda t},\qquad y''=\lambda^2Ye^{\lambda t}.

代入原方程組並消去 eλte^{\lambda t},得到

λ2X=−40X+8Y,\lambda^2X=-40X+8Y, λ2Y=12X−60Y.\lambda^2Y=12X-60Y.

整理為

(λ2+40)X−8Y=0,(\lambda^2+40)X-8Y=0, −12X+(λ2+60)Y=0.-12X+(\lambda^2+60)Y=0.

要有非零解 (X,Y)(X,Y),係數行列式必須為零:

∣λ2+40−8−12λ2+60∣=0.\begin{vmatrix} \lambda^2+40&-8\\ -12&\lambda^2+60 \end{vmatrix}=0.

因此

(λ2+40)(λ2+60)−96=0,(\lambda^2+40)(\lambda^2+60)-96=0, λ4+100λ2+2304=0.\lambda^4+100\lambda^2+2304=0.

因為

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 15 題

Continue from Question 14. Find the particular solution of the second order system in equation (2) with initial conditions x(0)=2,x′(0)=12,y(0)=1x(0) = 2, x'(0) = 12, y(0) = 1 and y′(0)=6y'(0) = 6. Which of the following statements is/are true?

(A) x(t)=2e6tx(t) = 2e^{6t}.
(B) x(t)=2cos⁡(6t)+2sin⁡(6t)x(t) = 2 \cos(6t) + 2 \sin(6t).
(C) y(t)=e6ty(t) = e^{6t}.
(D) y(t)=cos⁡(6t)+sin⁡(6t)y(t) = \cos(6t) + \sin(6t).
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查二階微分方程組的通解形式,以及初始條件的代入驗算。

對於特徵根:

  • 若特徵根為 λ=±6\lambda=\pm 6,解的形式為
    Ce6t+De−6t.Ce^{6t}+De^{-6t}.
  • 若特徵根為 λ=±6i\lambda=\pm 6i,解的形式為
    Ccos⁡6t+Dsin⁡6t.C\cos 6t+D\sin 6t.

本題的初始條件為

x(0)=2,x′(0)=12,x(0)=2,\qquad x'(0)=12,
y(0)=1,y′(0)=6.y(0)=1,\qquad y'(0)=6.

由於題目省略了第 14 題所引用的方程式(2),僅靠初始條件無法單獨判定應採用指數型或三角型解;必須依據方程式(2)的特徵根判斷。


解題方法

若第 14 題得到方程式(2)的特徵根為 λ=6\lambda=6,且由初始條件選出正向指數解,則

x(t)=2e6t,y(t)=e6t.x(t)=2e^{6t},\qquad y(t)=e^{6t}.

驗算初始條件:

x(0)=2e0=2,x(0)=2e^0=2,
x′(t)=12e6t  ⟹  x′(0)=12,x'(t)=12e^{6t}\implies x'(0)=12,

以及

y(0)=e0=1,y(0)=e^0=1,
y′(t)=6e6t  ⟹  y′(0)=6.y'(t)=6e^{6t}\implies y'(0)=6.

因此在方程式(2)確實對應正向指數模態時,選項 A、C 正確。

另一方面,若方程式(2)的特徵根為 λ=±6i\lambda=\pm 6i,則由初始條件可得

x(t)=2cos⁡(6t)+2sin⁡(6t),x(t)=2\cos(6t)+2\sin(6t),

因為

x(0)=2,x(0)=2,
x′(t)=−12sin⁡(6t)+12cos⁡(6t),x'(t)=-12\sin(6t)+12\cos(6t),

故

x′(0)=12.x'(0)=12.

同理,

y(t)=cos⁡(6t)+sin⁡(6t),y(t)=\cos(6t)+\sin(6t),

且

y′(t)=−6sin⁡(6t)+6cos⁡(6t),y'(t)=-6\sin(6t)+6\cos(6t),

所以

y(0)=1,y′(0)=6.y(0)=1,\qquad y'(0)=6.

因此 B、D 也都符合初始條件,但是否為方程式(2)的解,仍須檢查其是否符合第 14 題的微分方程組。


🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 16 題

Consider the following second order differential equation
3t2y′′(t)+sin⁡(t)y′(t)–cos⁡(t)y(t)=03t^2y''(t) + \sin(t)y'(t) – \cos(t)y(t) = 0.
Let y1(t)=tr1∑n=0∞antny_1(t) = t^{r_1} \sum_{n=0}^\infty a_n t^n and y2(t)=tr2∑n=0∞bntny_2(t) = t^{r_2} \sum_{n=0}^\infty b_n t^n be two linearly independent Frobenius series solutions for y(t)y(t) in the above differential equation when t>0t > 0. Assuming r1≥r2r_1 \ge r_2 and a0=1a_0 = 1, which of the following statements is/are true regarding the solution y1(t)y_1(t)?

(A) r1=1r_1 = 1.
(B) a1=12a_1 = \frac{1}{2}.
(C) a2=−160a_2 = -\frac{1}{60}.
(D) a3=11920a_3 = \frac{1}{1920}.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查:

  • Frobenius 級數法
  • 正則奇異點的指標方程
  • 將級數代回微分方程後比較各次方係數
  • 由遞迴關係求出 a1,a2,a3a_1,a_2,a_3

令

y1(t)=tr∑n=0∞antn,a0=1.y_1(t)=t^r\sum_{n=0}^{\infty}a_nt^n, \qquad a_0=1.

因為

sin⁡t=t−t36+⋯ ,cos⁡t=1−t22+⋯ ,\sin t=t-\frac{t^3}{6}+\cdots, \qquad \cos t=1-\frac{t^2}{2}+\cdots,

可先利用最低次方求出指標方程。


求指標根

取級數最低次項 y∼try\sim t^r,則

y′∼rtr−1,y′′∼r(r−1)tr−2.y'\sim rt^{r-1},\qquad y''\sim r(r-1)t^{r-2}.

代入

3t2y′′+sin⁡(t)y′−cos⁡(t)y=0,3t^2y''+\sin(t)y'-\cos(t)y=0,

取最低次 trt^r 項:

3r(r−1)tr+rtr−tr=0.3r(r-1)t^r+rt^r-t^r=0.

因此指標方程為

3r(r−1)+r−1=0,3r(r-1)+r-1=0,

即

3r2−2r−1=0.3r^2-2r-1=0.

因式分解得

(3r+1)(r−1)=0.(3r+1)(r-1)=0.

所以兩個指標根為

r1=1,r2=−13.r_1=1,\qquad r_2=-\frac13.

由題意 r1≥r2r_1\ge r_2,故

r1=1.r_1=1.

因此選項 (A) 正確。


解題方法:取 r1=1r_1=1 展開級數

此時

y1(t)=∑n=0∞antn+1=a0t+a1t2+a2t3+a3t4+⋯ .y_1(t)=\sum_{n=0}^{\infty}a_nt^{n+1} =a_0t+a_1t^2+a_2t^3+a_3t^4+\cdots.

因此

y1′(t)=a0+2a1t+3a2t2+4a3t3+⋯ ,y_1'(t)=a_0+2a_1t+3a_2t^2+4a_3t^3+\cdots, y1′′(t)=2a1+6a2t+12a3t2+⋯ .y_1''(t)=2a_1+6a_2t+12a_3t^2+\cdots.

又

sin⁡t=t−t36+⋯ ,cos⁡t=1−t22+⋯ .\sin t=t-\frac{t^3}{6}+\cdots, \qquad \cos t=1-\frac{t^2}{2}+\cdots.

將各項代入原方程:

3t2y1′′+sin⁡(t)y1′−cos⁡(t)y1=0.3t^2y_1''+\sin(t)y_1'-\cos(t)y_1=0.

求 a1a_1

比較 t2t^2 項:

  • 3t2y1′′3t^2y_1'' 提供 6a1t26a_1t^2
  • sin⁡(t)y1′\sin(t)y_1' 提供 2a1t22a_1t^2
  • −cos⁡(t)y1-\cos(t)y_1 提供 −a1t2-a_1t^2

所以

6a1+2a1−a1=0,6a_1+2a_1-a_1=0, 7a1=0.7a_1=0.

因此

a1=0.a_1=0.

選項 (B) 所稱 a1=12a_1=\frac12 錯誤。


求 a2a_2

比較 t3t^3 項:

由各部分可得

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 17 題

Continue from Question 16. Assuming b0=2b_0 = 2, which of the following statements is/are true regarding the solution y2(t)y_2(t)?

(A) r2=−13r_2 = -\frac{1}{3}.
(B) b1=0b_1 = 0.
(C) b2=−518b_2 = -\frac{5}{18}.
(D) b3=6112960b_3 = \frac{61}{12960}.
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題承接第 16 題微分方程

3t2y′′(t)+sin⁡(t)y′(t)−cos⁡(t)y(t)=0,3t^2y''(t)+\sin(t)y'(t)-\cos(t)y(t)=0,

要求利用 Frobenius 級數法,求第二組解

y2(t)=∑n=0∞bntn+r2.y_2(t)=\sum_{n=0}^{\infty}b_nt^{n+r_2}.

主要觀念包括:

  • Frobenius 解的形式;
  • 指標方程(indicial equation);
  • 由同次方冪係數比較得到遞迴關係;
  • 逐項求出 r2,b1,b2,b3r_2,b_1,b_2,b_3。

解題方法

設

y2(t)=∑n=0∞bntn+r,b0≠0.y_2(t)=\sum_{n=0}^{\infty}b_nt^{n+r}, \qquad b_0\neq 0.

則

y2′(t)=∑n=0∞(n+r)bntn+r−1,y_2'(t)=\sum_{n=0}^{\infty}(n+r)b_nt^{n+r-1}, y2′′(t)=∑n=0∞(n+r)(n+r−1)bntn+r−2.y_2''(t)=\sum_{n=0}^{\infty}(n+r)(n+r-1)b_nt^{n+r-2}.

又有

sin⁡t=t−t33!+t55!−⋯ ,\sin t=t-\frac{t^3}{3!}+\frac{t^5}{5!}-\cdots, cos⁡t=1−t22!+t44!−⋯ .\cos t=1-\frac{t^2}{2!}+\frac{t^4}{4!}-\cdots.

代入微分方程後,最低次 trt^r 的係數為

[3r(r−1)+r−1]b0=0.\left[3r(r-1)+r-1\right]b_0=0.

因為 b0≠0b_0\neq 0,得到指標方程

3r(r−1)+r−1=0,3r(r-1)+r-1=0,

即

3r2−2r−1=0=(3r+1)(r−1).3r^2-2r-1=0 =(3r+1)(r-1).

因此兩個指標根為

r1=1,r2=−13.r_1=1,\qquad r_2=-\frac13.

所以選項(A)正確。


求係數遞迴關係

對於 tn+rt^{n+r} 的係數,來自主要項的係數為

[3(n+r)(n+r−1)+(n+r)−1]bn.\left[3(n+r)(n+r-1)+(n+r)-1\right]b_n.

由 sin⁡t\sin t 與 cos⁡t\cos t 的高次項,可得一般遞迴式

[3(n+r)(n+r−1)+(n+r)−1]bn+∑k=1⌊n/2⌋(−1)k(2k+1)!(n−2k+r−(2k+1))bn−2k=0.\begin{aligned} &\left[3(n+r)(n+r-1)+(n+r)-1\right]b_n\\ &\quad+\sum_{k=1}^{\lfloor n/2\rfloor} \frac{(-1)^k}{(2k+1)!} \left(n-2k+r-(2k+1)\right)b_{n-2k}=0. \end{aligned}

實際計算前幾項即可判斷所有選項。


計算 b1b_1

令 r=r2=−13r=r_2=-\frac13,取 n=1n=1。因為沒有較低階的前項,故

[3(1−13)(1−13−1)+(1−13)−1]b1=0.\left[3\left(1-\frac13\right)\left(1-\frac13-1\right) +\left(1-\frac13\right)-1\right]b_1=0.

計算括號:

3(23)(−13)+23−1=−23−13=−1.3\left(\frac23\right)\left(-\frac13\right) +\frac23-1 =-\frac23-\frac13=-1.

因此

−b1=0⟹b1=0.-b_1=0 \quad\Longrightarrow\quad b_1=0.

所以選項(B)正確。

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊
📄 以下 2 題共用同一段題幹

Consider the following differential equation for the function y(t)y(t) defined for t≥0t\geq 0:
y′(t)=2t[y(t)⋆(cosh⁡(t)u(t))].y'(t)=\frac{2}{t}\left[y(t)\star\left(\cosh(t)u(t)\right)\right].

Here u(t)u(t) is the usual unit-step (Heaviside) function and ⋆\star is the usual convolutional operation of two functions. Assume y(t)=0y(t)=0 for t<0t<0. The initial condition is y(0)=1y(0)=1.

第 18 題

Which of the following statements is/are true regarding the Laplace transform Y(s)=L{y(t)}Y(s)=\mathcal{L}\{y(t)\} of solution y(t)y(t) in the above equation?

(A) Y(s=12)=−6Y\left(s=\frac{1}{2}\right)=-6
(B) Y(s=1)=1Y(s=1)=1
(C) Y(s=2)=32Y(s=2)=\frac{3}{2}
(D) Y(s=3)=43Y(s=3)=\frac{4}{3}
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考查**拉氏轉換(Laplace Transform)**求解變係數微分積分方程式,主要運用以下定義與定理:

  1. 頻域微分性質(乘 tt 性質):
    L{tf(t)}=−ddsF(s)\mathcal{L}\{t f(t)\} = -\frac{d}{ds}F(s)
  2. 一階微分性質:
    L{y′(t)}=sY(s)−y(0)\mathcal{L}\{y'(t)\} = s Y(s) - y(0)
  3. 摺積定理(Convolution Theorem):
    L{f(t)⋆g(t)}=F(s)G(s)\mathcal{L}\{f(t) \star g(t)\} = F(s) G(s)
  4. 雙曲餘弦函數的拉氏轉換:
    L{cosh⁡(t)u(t)}=ss2−1(Re⁡(s)>1)\mathcal{L}\{\cosh(t)u(t)\} = \frac{s}{s^2 - 1} \quad (\operatorname{Re}(s) > 1)
  5. 初值定理(Initial Value Theorem):
    lim⁡s→∞sY(s)=y(0+)\lim_{s \to \infty} s Y(s) = y(0^+)
  6. 一階常微分方程式分離變數法與**拉氏轉換之收斂域(ROC)**概念。

解題方法

原卷第 14 頁第 18 題給定因果信號微分方程式:
y′(t)=2t[y(t)⋆(cosh⁡(t)u(t))],t≥0y'(t) = \frac{2}{t} \left[ y(t) \star (\cosh(t)u(t)) \right], \quad t \ge 0
初始條件為 y(0)=1y(0) = 1,y(t)=0y(t) = 0 對於 t<0t < 0。

步驟一:去分母並取拉氏轉換

將兩邊同乘以 tt:
ty′(t)=2[y(t)⋆(cosh⁡(t)u(t))]t y'(t) = 2 \left[ y(t) \star (\cosh(t)u(t)) \right]

對等號兩端取拉氏轉換:

  • 左端利用乘 tt 性質與微分性質(已知 y(0)=1y(0) = 1):
    L{ty′(t)}=−ddsL{y′(t)}=−dds[sY(s)−y(0)]=−dds[sY(s)]=−[Y(s)+sY′(s)]\mathcal{L}\{t y'(t)\} = -\frac{d}{ds}\mathcal{L}\{y'(t)\} = -\frac{d}{ds}\left[ s Y(s) - y(0) \right] = -\frac{d}{ds}\left[ s Y(s) \right] = -[Y(s) + s Y'(s)]
  • 右端利用摺積定理:
    L{2[y(t)⋆(cosh⁡(t)u(t))]}=2Y(s)⋅ss2−1=2ss2−1Y(s)\mathcal{L}\left\{ 2 [y(t) \star (\cosh(t)u(t))] \right\} = 2 Y(s) \cdot \frac{s}{s^2 - 1} = \frac{2s}{s^2 - 1} Y(s)

步驟二:建立並求解 Y(s)Y(s) 的微分方程式

兩端相等得:
−[Y(s)+sY′(s)]=2ss2−1Y(s)-[Y(s) + s Y'(s)] = \frac{2s}{s^2 - 1} Y(s)
整理項次:
−sY′(s)=(1+2ss2−1)Y(s)-s Y'(s) = \left( 1 + \frac{2s}{s^2 - 1} \right) Y(s)
兩端同除以 −sY(s)-s Y(s),分離變數:
Y′(s)Y(s)=−1s−2s2−1\frac{Y'(s)}{Y(s)} = -\frac{1}{s} - \frac{2}{s^2 - 1}

利用部分分式拆解第二項:
2s2−1=1s−1−1s+1\frac{2}{s^2 - 1} = \frac{1}{s - 1} - \frac{1}{s + 1}
因此微分式為:
Y′(s)Y(s)=−1s−1s−1+1s+1\frac{Y'(s)}{Y(s)} = -\frac{1}{s} - \frac{1}{s - 1} + \frac{1}{s + 1}

兩端對 ss 積分:
ln⁡Y(s)=−ln⁡s−ln⁡(s−1)+ln⁡(s+1)+C1=ln⁡(s+1s(s−1))+C1\ln Y(s) = -\ln s - \ln(s - 1) + \ln(s + 1) + C_1 = \ln \left( \frac{s + 1}{s(s - 1)} \right) + C_1
取指數得:
Y(s)=Cs+1s(s−1)Y(s) = C \frac{s + 1}{s(s - 1)}

🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 19 題

Continue from Question 十八. Which of the following statements is/are true regarding the solution y(t)y(t)?

(A) y′(1)=1y'(1)=1
(B) y′(π)=1y'(\pi)=1
(C) y′′(1)=−1y''(1)=-1
(D) y′′(π)=0y''(\pi)=0
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

本題考拉普拉斯轉換中的卷積定理,以及 tf(t)t f(t) 的轉換公式:

L{f∗g}=F(s)G(s),L{tf(t)}=−ddsF(s).\mathcal{L}\{f*g\}=F(s)G(s),\qquad \mathcal{L}\{t f(t)\}=-\frac{d}{ds}F(s).

由圖可讀到共用方程為 y′(t)=2t[y(t)⋆(cosh⁡(t)u(t))]y'(t)=\frac{2}{t}[y(t)\star(\cosh(t)u(t))],並給定 y(t)=0y(t)=0(t<0t<0)及 y(0)=1y(0)=1;第 19 題詢問 y′(1)y'(1)、y′(π)y'(\pi)、y′′(1)y''(1)、y′′(π)y''(\pi) 的值。

解題方法

令 Y(s)=L{y(t)}Y(s)=\mathcal{L}\{y(t)\}。先將方程乘以 tt:

ty′(t)=2[y(t)⋆(cosh⁡(t)u(t))].t y'(t)=2\bigl[y(t)\star(\cosh(t)u(t))\bigr].

因為 L{y′(t)}=sY(s)−y(0)=sY(s)−1\mathcal{L}\{y'(t)\}=sY(s)-y(0)=sY(s)-1,所以

L{ty′(t)}=−dds[sY(s)−1]=−Y(s)−sY′(s).\mathcal{L}\{t y'(t)\} =-\frac{d}{ds}\bigl[sY(s)-1\bigr] =-Y(s)-sY'(s).

又 L{cosh⁡(t)u(t)}=ss2−1\mathcal{L}\{\cosh(t)u(t)\}=\frac{s}{s^2-1}。利用卷積定理,得到

−Y−sY′=2Yss2−1,-Y-sY'=2Y\frac{s}{s^2-1},

整理為

Y′Y=−1s−2s2−1.\frac{Y'}{Y}=-\frac{1}{s}-\frac{2}{s^2-1}.
🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

第 20 題

Consider the following partial differential equation for the bivariate function v(x,t)v(x,t):
∂∂tv(x,t)=∂2∂x2v(x,t),0≤x≤10,t≥0,\frac{\partial}{\partial t}v(x,t)=\frac{\partial^2}{\partial x^2}v(x,t),\qquad 0\leq x\leq 10,\quad t\geq 0,
subject to conditions
∂∂xv(x,t)∣x=0=∂∂xv(x,t)∣x=5=0,for all t≥0,\left.\frac{\partial}{\partial x}v(x,t)\right|_{x=0}=\left.\frac{\partial}{\partial x}v(x,t)\right|_{x=5}=0,\qquad \text{for all }t\geq 0,
v(x,0)=x2−10x,for all x with 0<x<10.v(x,0)=x^2-10x,\qquad \text{for all }x\text{ with }0<x<10.

Express the solution v(x,t)v(x,t) in the following form:
v(x,t)=∑n=0∞[ancos⁡(pnπx)+bnsin⁡(qnπx)]exp⁡(−rnπ2t),v(x,t)=\sum_{n=0}^{\infty}\left[a_n\cos(p_n\pi x)+b_n\sin(q_n\pi x)\right]\exp(-r_n\pi^2t),
for some constants an,bn,pn,qn,rn∈Ra_n,b_n,p_n,q_n,r_n\in\mathbb{R} satisfying 0≤r0<r1<r2<⋯0\leq r_0<r_1<r_2<\cdots and an2+bn2>0a_n^2+b_n^2>0 for n=1,2,…n=1,2,\ldots. Which of the following statements is/are true?

(A) a0=−503a_0=-\frac{50}{3}
(B) b1=1729b_1=\frac{17}{29}
(C) p3=910p_3=\frac{9}{10}
(D) r4=3250r_4=\frac{32}{50}
(E) None of the above is true.

登入後即可作答並保存紀錄。

這一題的完整詳解

核心觀念

這題考熱方程的分離變數與傅立葉餘弦級數。對 0≤x≤100\le x\le 10 的初始函數作餘弦展開,再利用初始函數關於 x=5x=5 對稱的性質,找出哪些模態會留下。

解題方法

依原卷,空間範圍是 0≤x≤100\le x\le 10,邊界條件為 vx(0,t)=vx(5,t)=0v_x(0,t)=v_x(5,t)=0,初始條件為 v(x,0)=x2−10xv(x,0)=x^2-10x。初始函數滿足

(10−x)2−10(10−x)=x2−10x,(10-x)^2-10(10-x)=x^2-10x,

所以其餘弦展開只含全區間標準模態中的偶數項,也就是 cos⁡(nπx/5)\cos(n\pi x/5)。

令

v(x,t)=a0+∑n=1∞ancos⁡(nπx5)exp⁡(−n2π225t).v(x,t)=a_0+\sum_{n=1}^{\infty}a_n\cos\left(\frac{n\pi x}{5}\right) \exp\left(-\frac{n^2\pi^2}{25}t\right).

這些模態在 x=0x=0 與 x=5x=5 的空間導數皆為零,符合圖中的邊界條件。

常數項是初始函數在 [0,10][0,10] 上的平均值:

a0=110∫010(x2−10x) dx=−503.a_0=\frac{1}{10}\int_0^{10}(x^2-10x)\,dx=-\frac{50}{3}.
🔒

後續完整解題步驟與【答案】

免費註冊,享三天全站完整詳解閱覽。

免費註冊

其他考古題