I1, I2 - Integrating factors

I1, I2 - Integrating factors

An equation can tell you how a quantity changes without giving the quantity itself. I1 introduces a multiplier that turns a first-order linear differential equation into a single product derivative; I2 uses it to find a whole family of solutions. You will recognise the right form, prepare the equation and explain why the method works.

Recognising a linear first-order equation

How can we find all the functions satisfying y′+2y=6xy'+2y=6x, when neither term on the left can simply be ignored? Here y′=dy/dxy'=\mathrm{d}y/\mathrm{d}x: the unknown is a function, not a number.

Integrating term by term would give y+2∫y dx=3x2+Cy+2\int y\,\mathrm{d}x=3x^2+C, which still contains an unknown integral. Instead, we will make the left side the derivative of a product.

Linear first-order differential equation

For an unknown function y(x)y(x), a linear first-order differential equation can be written

dydx+P(x)y=Q(x)\frac{\mathrm{d}y}{\mathrm{d}x}+P(x)y=Q(x)

on an interval where its coefficients are defined. The highest derivative is the first derivative, and yy and y′y' occur only to the first power, without being multiplied together. The coefficients PP and QQ depend only on the independent variable xx.

For y′+2y=6xy'+2y=6x, P(x)=2P(x)=2 and Q(x)=6xQ(x)=6x. “Linear” describes dependence on yy, not on xx: y′+x2y=sin⁡xy'+x^2y=\sin x is still linear. A missing yy-term means P=0P=0, and a zero right side means Q=0Q=0; both are allowed.

By contrast, y′+y2=xy'+y^2=x is nonlinear because of y2y^2, and yy′+y=xyy'+y=x is nonlinear because the unknown multiplies its derivative. The equation y′′+y=xy''+y=x is linear but second order. These labels matter because a method has conditions, not just a familiar-looking symbol.

Key idea

Look at how the unknown function and its derivatives appear. A coefficient such as x2x^2 is allowed; a coefficient involving the unknown yy is not part of this linear form.

Try that distinction before doing any integration.

Making the derivative coefficient one

The coefficient used in an integrating factor is the coefficient of yy after the coefficient of y′y' has been made one.

If the equation is a(x)y′+b(x)y=c(x)a(x)y'+b(x)y=c(x), divide every term by a(x)a(x). This is valid only where a(x)≠0a(x)\ne0. Work on a connected interval avoiding any excluded point; a solution on one side of such a point does not automatically extend through it.

Worked example 1: a constant coefficient to divide out

Question. Put 2y′+4y=6x2y'+4y=6x in standard linear form and identify PP and QQ.

Plan. Divide every term by the coefficient of y′y'.

2y′+4y=6x⟹y′+2y=3x2y'+4y=6x \quad\Longrightarrow\quad y'+2y=3x

The divisor is the nonzero constant 22, so it introduces no excluded value of xx.

Answer. P(x)=2P(x)=2 and Q(x)=3xQ(x)=3x, for all real xx.

Check. Multiplying the standard form by 22 recovers all three original terms.

Now change only the derivative coefficient from 22 to 2x2x. Division changes both the coefficient of yy and the allowed interval.

Worked example 2: a variable coefficient to divide out

Question. Put 2xy′+4y=6x2xy'+4y=6x in standard linear form, stating its domain restriction.

Plan. Divide by 2x2x, keeping the restriction next to the result.

2xy′+4y=6xy′+2xy=3(x≠0)\begin{aligned} 2xy'+4y&=6x\\ y'+\frac{2}{x}y&=3 && (x\ne0) \end{aligned}

Answer. P(x)=2/xP(x)=2/x, Q(x)=3Q(x)=3. Use the standard form on x>0x>0 or on x<0x<0, separately.

Check. Multiplication by 2x2x returns the original equation when x≠0x\ne0. At x=0x=0, the original equation instead requires 4y(0)=04y(0)=0; division has not analysed whether any solution can pass through that point.

The original equation may contain information at an excluded point, but it needs separate examination there.

Watch out

Wrong: reading P=4P=4 directly from 2xy′+4y=6x2xy'+4y=6x. After normalising, P=2/xP=2/x, not 44. Divide the right side as well.

Prepare a new equation in the same way.

Deriving the integrating factor

For our opening equation, multiplying by e2xe^{2x} creates e2xy′+2e2xye^{2x}y'+2e^{2x}y. The product rule says this is exactly (e2xy)′(e^{2x}y)'. The multiplier has made the extra product-rule term useful.

For a general multiplier μ(x)\mu(x), compare

μy′+μPy=μQ,(μy)′=μy′+μ′y\mu y'+\mu Py=\mu Q, \qquad (\mu y)'=\mu y'+\mu' y

We need μ′=Pμ\mu'=P\mu. A convenient nonzero solution is obtained from μ′/μ=P\mu'/\mu=P.

Integrating factor

For y′+P(x)y=Q(x)y'+P(x)y=Q(x), with continuous P,QP,Q on the interval being considered, choose

μ(x)=e∫P(x) dx\mu(x)=e^{\int P(x)\,\mathrm{d}x}

Then μ′=Pμ\mu'=P\mu, so

ddx(μy)=μQ\frac{\mathrm{d}}{\mathrm{d}x}(\mu y)=\mu Q

Not in the formula booklet: learn it.

Here the integral means any one antiderivative. Adding a constant would only multiply μ\mu by a nonzero constant; this scales both sides and gives the same solutions. We therefore omit that unnecessary constant when finding the factor.

Continue this lesson.

You’re previewing NovaLearn without an account. Create a free account to read the rest.

Create free account