You need to enable JavaScript to run this app.
优惠活动
大模型
产品
解决方案
定价
更多

如何利用已证矩阵指数求解非齐次矩阵微分方程(含初始条件)

Hey there! Let's work through this step by step since you already have the matrix exponential $e^{Pu}$ handy—half the battle is already won here. 😊

Step 1: Use the constant variation formula for linear nonhomogeneous ODEs

For a linear nonhomogeneous system of the form:
$$\frac{d}{du}\mathbf{z}(u) = P\mathbf{z}(u) + \mathbf{f}(u)$$
with initial condition $\mathbf{z}(0) = \mathbf{z}_0$, the exact solution is given by the constant variation formula:
$$
\mathbf{z}(u) = e^{Pu}\mathbf{z}_0 + \int_0^u e^{P(u-s)}\mathbf{f}(s)ds
$$
Since your initial condition is $\mathbf{z}(0) = \begin{bmatrix}0\0\end{bmatrix}$, the first term $e^{Pu}\mathbf{z}_0$ vanishes entirely. We only need to compute the integral term.

Step 2: Substitute the given matrix exponential into the integral

First, rewrite $e^{P(u-s)}$ using the provided $e^{Pt}$ form. Replace $t$ with $(u-s)$:
$$
e^{P(u-s)} = e^{-x(u-s)} \begin{bmatrix}\cos(y(u-s)) & \sin(y(u-s))\ -\sin(y(u-s)) & \cos(y(u-s))\end{bmatrix}
$$
Your forcing term is $\mathbf{f}(s) = \begin{bmatrix}\cos(zs)\ -\sin(zs)\end{bmatrix}$. Compute the matrix-vector product $e^{P(u-s)}\mathbf{f}(s)$:

First component (for $a(u)$):

$$
e^{-x(u-s)} \left[ \cos(y(u-s))\cos(zs) - \sin(y(u-s))\sin(zs) \right]
$$
Use the cosine addition identity $\cos(A+B) = \cos A \cos B - \sin A \sin B$ (here $A = y(u-s)$, $B = zs$) to simplify this to:
$$
e^{-x(u-s)} \cos\left( yu + s(z - y) \right)
$$

Second component (for $b(u)$):

$$
e^{-x(u-s)} \left[ -\sin(y(u-s))\cos(zs) - \cos(y(u-s))\sin(zs) \right]
$$
Use the sine addition identity $\sin(A+B) = \sin A \cos B + \cos A \sin B$, then factor out a negative sign:
$$
-e^{-x(u-s)} \sin\left( yu + s(z - y) \right)
$$

Step 3: Evaluate the integrals for $a(u)$ and $b(u)$

Now we compute the definite integrals for each component. Let's start with $a(u)$:

Calculating $a(u)$:

$$
a(u) = \int_0^u e^{-x(u-s)} \cos\left( yu + s(z - y) \right) ds
$$
Make a substitution $t = u - s$ (so $ds = -dt$, bounds swap from $t=u$ to $t=0$):
$$
a(u) = e^{-xu} \int_0^u e^{xt} \cos\left( zu - t(z - y) \right) dt
$$
Expand the cosine term using $\cos(C-D) = \cos C \cos D + \sin C \sin D$:
$$
a(u) = e^{-xu} \left[ \cos(zu) \int_0^u e^{xt} \cos\left( t(z-y) \right) dt + \sin(zu) \int_0^u e^{xt} \sin\left( t(z-y) \right) dt \right]
$$
Use standard exponential-trigonometric integral formulas:

  • $\int e^{kt} \cos(mt) dt = \frac{e{kt}}{k2+m^2}\left(k\cos(mt)+m\sin(mt)\right) + C$
  • $\int e^{kt} \sin(mt) dt = \frac{e{kt}}{k2+m^2}\left(k\sin(mt)-m\cos(mt)\right) + C$

Substitute $k=x$, $m=z-y$, evaluate from $0$ to $u$, then simplify the expression.

Calculating $b(u)$:

$$
b(u) = -\int_0^u e^{-x(u-s)} \sin\left( yu + s(z - y) \right) ds
$$
Follow the same substitution $t=u-s$, expand using $\sin(C-D) = \sin C \cos D - \cos C \sin D$, and apply the same integral formulas to get a closed-form solution.

Step 4: Simplify the final results

After evaluating the integrals, you can distribute the $e^{-xu}$ term and combine like terms to get a clean, closed-form expression for both $a(u)$ and $b(u)$. For example, the simplified $a(u)$ might look like:
$$
a(u) = \frac{1}{x2+(z-y)2}\left[ x\cos(zu)+(z-y)\sin(zu) - e^{-xu}\left(x\cos(yu)+(z-y)\sin(yu)\right) \right]
$$


内容的提问来源于stack exchange,提问作者Hideki Ryuga

相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级,最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分), 限时 9.9 元起

最近更新时间:2026.05.19 09:16:44