Differential and Integral Calculus

Introduction

Differential and integral calculus provides the mathematical language for change and accumulation. In electronics and communication engineering, derivatives describe signal slope, current–voltage sensitivity, transient response and optimization. Integrals determine electric charge, energy, average signal value and accumulated error. Limits establish the foundations of both operations, while sequences, series and numerical methods enable practical approximation when closed-form solutions are unavailable.

This chapter develops these ideas with attention to theorem conditions, convergence, numerical error and dimensional consistency.

Learning Objectives

After studying this chapter, students should be able to:

1. Determine the convergence or divergence of basic sequences.
2. Evaluate and interpret limits and continuity.
3. Compute derivatives and use them to analyze engineering rates and extrema.
4. construct Taylor and Maclaurin approximations with valid remainder conditions.
5. Evaluate integrals using substitution, integration by parts and other standard methods.
6. Analyze definite and improper integrals and their convergence.
7. Apply numerical differentiation and quadrature with quantitative error estimates.
8. Calculate accumulated charge, energy and related engineering quantities.

1. Sequences and Convergence

A sequence is an ordered list of real or complex numbers, written \{a_n\}_n=1^\infty. It is formally a function whose domain is the positive integers.

1.1 Definition of convergence

A sequence \{a_n\} converges to a finite limit L if its terms can be made arbitrarily close to L by taking n sufficiently large. Formally,

means that for every \varepsilon>0, there exists an integer N such that

A sequence that has no finite limit is divergent.

Example: For a_n=1/n, choose N>1/\varepsilon. Then n\ge N implies |a_n|=1/n<\varepsilon. Thus a_n\to0.

A geometric sequence a_n=arⁿ converges to zero when |r|<1. It generally diverges when |r|>1. For r=1, it is constant, while r=-1 produces oscillation unless a=0.

1.2 Monotone convergence and Cauchy sequences

A sequence is increasing if a_n+1\ge a_n and decreasing if a_n+1\le a_n. Every monotone sequence that is bounded converges.

A real or complex sequence is convergent exactly when it is a Cauchy sequence: for every \varepsilon>0, there exists N such that

This criterion is useful when the limit is not known in advance. In numerical algorithms, successive approximations often form a sequence; convergence means that increasing the iteration count approaches a stable mathematical value.

2. Limits and Continuity of Functions

2.1 Function limits

The statement

means that f(x) approaches L as x approaches a, whether or not f(a) is defined. Formally, for every \varepsilon>0, there exists \delta>0 such that

The two-sided limit exists only if the left-hand and right-hand limits exist and are equal. Standard limit laws permit sums, products, quotients with nonzero denominator limits and compositions with continuous functions.

2.2 Indeterminate forms

Expressions such as 0/0 and \infty/\infty are indeterminate forms, not numerical values. Algebraic simplification, standard limits or l’Hôpital’s rule may resolve them. If f and g are differentiable near a, g'(x)\ne0, and the required derivative-ratio limit exists, then under the hypotheses of l’Hôpital’s rule,

For example,

Angles must be measured in radians for this limit and the standard trigonometric derivative formulas.

2.3 Continuity

A function f is continuous at x=a if:

1. f(a) is defined;
2. \lim_{x\to a}f(x) exists;
3. \lim_{x\to a}f(x)=f(a).

Continuity on an interval means continuity at every interior point, together with appropriate one-sided continuity at endpoints.

The Intermediate Value Theorem states that if f is continuous on [a,b] and y lies between f(a) and f(b), then some c\in[a,b] satisfies f(c)=y. This theorem supports numerical root-finding methods such as bisection.

A signal model with a jump discontinuity can still be physically useful, but its derivative does not exist at the jump in the ordinary sense. Real hardware also has finite bandwidth, so ideal discontinuities are approximations.

3. Derivatives and Rates of Change

3.1 Definition and interpretation

The derivative of f at x is

provided the finite limit exists. Geometrically, it is the tangent slope. Physically, it is an instantaneous rate.

If voltage v(t) is measured in volts and time t in seconds, then dv/dt has units V·s⁻¹. Differentiation changes units; this must be respected in engineering equations.

3.2 Differentiation rules

For differentiable functions u(x) and v(x) and constant c,

For a composite function, the chain rule is

If an equation F(x,y)=0 defines y implicitly and F_y\ne0, implicit differentiation gives

3.3 Higher derivatives and extrema

The second derivative f''(x) measures the rate of change of slope. If f'(c)=0 and f''(c)>0, then c is a strict local minimum; if f''(c)<0, it is a strict local maximum. If f''(c)=0, the test is inconclusive.

The Mean Value Theorem states that if f is continuous on [a,b] and differentiable on (a,b), then some c\in(a,b) satisfies

It connects average and instantaneous rates and underlies error estimates in approximation theory.

3.4 Worked engineering example: RC charging

For an ideal series RC circuit driven by a DC step V_s at t=0, with initially uncharged capacitor, the capacitor voltage is

Here R is resistance in ohms, C is capacitance in farads and RC is the time constant in seconds. Differentiation gives

Because capacitor current is i_C=C\,dv_C/dt,

At t=0, the current is V_s/R, and it approaches zero as t\to\infty. These conclusions agree with the physical boundary conditions of an initially uncharged capacitor and its eventual steady state.

4. Linearization, Taylor Polynomials and Series

4.1 Local linear approximation

Near x=a, a differentiable function may be approximated by its tangent line:

This approximation is useful for small-signal circuit analysis, where a nonlinear device characteristic is linearized about an operating point.

4.2 Taylor’s theorem

Suppose f has the required derivatives through order n+1 on an interval containing a and x. Taylor’s theorem gives the exact finite identity

where the Lagrange remainder is

for some \xi between a and x, under the standard continuity and differentiability hypotheses of Taylor’s theorem.

The Taylor polynomial is only the finite sum. The associated infinite Taylor series is

A function equals this infinite series only at points where R_n(x)\to0 as n\to\infty. Equivalently, equality is guaranteed when the function is analytic on the relevant neighborhood. Infinite differentiability alone is not sufficient.

For example, define

This function is infinitely differentiable and every derivative at zero equals zero. Its Maclaurin series is therefore identically zero, yet g(x)>0 for x\ne0. Thus it is smooth but not represented by its Taylor series at zero.

4.3 Common Maclaurin series

A Maclaurin series is a Taylor series centered at a=0.

4.4 Worked approximation with an error bound

Approximate e^0.2 using the cubic Maclaurin polynomial:

At x=0.2,

Since the fourth derivative of e^x is e^x, the remainder satisfies

The actual value is approximately 1.22140276, giving an absolute error of about 6.94\times10^-5, within the bound.

5. Numerical Differentiation

Measured signals are usually available only at discrete sample times. If samples are separated by h, the forward-difference approximation is

If f'' is continuous, its leading truncation error is proportional to h; hence it is first-order accurate. More precisely, for some \xi\in(x,x+h),

The central difference is

If f''' is continuous, then for an appropriate intermediate point \xi,

It is therefore second-order accurate.

Worked calculation

For f(x)=\sin x, estimate f'(0.5) using h=0.1 radians:

The exact derivative is \cos(0.5)=0.87758256, so the absolute error is approximately 0.00146191. Because |f'''(x)|=|{-\cos x}|\le1, the theoretical error bound is

The computed error satisfies the bound. Very small h does not always improve a floating-point result because subtraction and measurement noise can amplify round-off error.

6. Integration and Accumulation

6.1 Definite integrals and the Fundamental Theorem

For an integrable function f, the definite integral

is the limit of suitable Riemann sums and represents signed accumulation. If F'(x)=f(x) on [a,b], the Fundamental Theorem of Calculus gives

If current i(t) is measured in amperes and time in seconds, then

has units A·s, or coulombs. If instantaneous power p(t) is in watts, \int p(t)dt is energy in joules.

The average value of an integrable function over [a,b] is

6.2 Substitution

Substitution reverses the chain rule. If u=g(x) and the hypotheses for change of variables are satisfied, then

Worked example:

Let u=x², so du=2x\,dx. The limits remain 0 and 1:

6.3 Integration by parts

The product rule yields

For a>0 and T\ge0, consider

Take u=t and dv=e^-atdt. Then du=dt and v=-e^-at/a, giving

This form occurs in transient-response moment calculations. The parameter a has units of inverse time, making the result have units of time squared.

6.4 Partial fractions

Rational functions can often be integrated after decomposition. For example,

so, on intervals not crossing x=0 or x=-1,

7. Improper Integrals and Convergence

An integral is improper when the interval is unbounded or the integrand becomes unbounded. It is defined through a limit.

For example,

provided the limit exists and is finite. Otherwise, the integral diverges.

The standard p-integrals are

and

For a singularity at an interior point c, both integrals \int_a^c f(x)dx and \int_c^b f(x)dx must converge separately. Infinite cancellations cannot be used to declare ordinary improper convergence.

For nonnegative functions, comparison is useful. If 0\le f(x)\le g(x) for sufficiently large x and \int g converges, then \int f converges. Conversely, if f(x)\ge g(x)\ge0 and \int g diverges, then \int f diverges.

8. Numerical Integration

Let h=(b-a)/n and x_j=a+jh. The composite trapezoidal rule is

If f'' is continuous, then for some \xi\in(a,b),

Thus its global truncation error is O(h²).

For even n, composite Simpson’s rule is

If f⁽⁴⁾ is continuous, then for some \xi\in(a,b),

Simpson’s rule is fourth-order accurate and is exact for polynomials of degree at most three.

Worked comparison

Approximate I=\int₀¹e^-xdx with n=4 and h=0.25. The sampled values are:

The trapezoidal estimate is

Simpson’s estimate is

The exact value is 1-e^-1=0.632121. The absolute errors are approximately 3.289\times10^-3 and 1.36\times10^-5, respectively.

Because f⁽⁴⁾(x)=e^-x\le1 on [0,1], Simpson’s error is bounded by

9. Illustrative Case: Charge and Energy from a Pulsed Current

Background and problem

This fictional engineering case concerns a current driver tested over 0\le t\le2\ \mathrm{s}. Its measured behavior is modeled by

The coefficient includes units so that current is measured in milliamperes. The terminal voltage is modeled by

where V₀=5\ \mathrm{V} and \alpha=2\ \mathrm{V\,s^-1}. Determine transferred charge and supplied energy.

Analysis and solution

The charge is

Instantaneous power is p(t)=v(t)i(t). Since V×mA=mW,

where the displayed numerical polynomial assumes t is inserted in seconds with the stated coefficient units. Hence

Using samples every 0.5\ \mathrm{s}, the currents are 0, 1, 4, 9 and 16 mA. The trapezoidal estimate of charge is 11 mC. Simpson’s rule gives

Simpson’s result is exact here because current is quadratic. It is also exact for the cubic power model, producing E=85.3333\ \mathrm{mJ}.

Outcome and learning points

The case demonstrates that derivatives and integrals must retain physical units, numerical accuracy depends on the mathematical form of the signal, and sample spacing affects accumulated-quantity estimates. In real testing, sensor calibration, bandwidth, noise and model validity must also be assessed.

10. Practical Limitations and Professional Practice

Calculus models are only as reliable as their assumptions. Ideal component laws may fail under saturation, temperature variation, parasitic effects or high-frequency operation. Numerical differentiation strongly amplifies measurement noise, while numerical integration can accumulate sensor bias.

Good engineering practice includes:

• recording variables and units explicitly;
• checking continuity and differentiability assumptions;
• reporting approximation order and error estimates;
• refining the step size and checking convergence;
• comparing numerical results with physical bounds;
• retaining raw data and documenting preprocessing;
• independently validating software or AI-generated calculations.

Computer algebra and AI tools can assist with symbolic differentiation, integration and error analysis, but they may silently omit convergence conditions or choose inappropriate branches. Professional responsibility remains with the engineer, especially where computed energy, current or safety limits influence hardware decisions or regulatory compliance.

Chapter Summary

Sequences provide a foundation for convergence and numerical approximation. Limits define continuity and derivatives, while derivatives describe instantaneous rates, sensitivity and extrema. Taylor polynomials provide controlled local approximations, but a function equals its infinite Taylor series only where the remainder tends to zero; smoothness alone is insufficient.

Definite integrals represent accumulated quantities and can be evaluated through antiderivatives, substitution, integration by parts and partial fractions. Improper integrals require convergence analysis through limits. Numerical differentiation and integration enable calculations from sampled data, but their truncation errors, round-off effects and physical units must be evaluated. These tools support reliable analysis of signal slopes, transient responses, charge and energy in electronic systems.

Key Terms

Sequence: An ordered function of the positive integers.

Convergence: Approach to a finite limiting value according to a precise error criterion.

Cauchy sequence: A sequence whose sufficiently late terms are arbitrarily close to one another.

Limit: The value approached by a function or sequence.

Continuity: The property that a function value equals its limit at a point.

Derivative: The limiting instantaneous rate of change of a function.

Chain rule: A rule for differentiating composite functions.

Critical point: A domain point where the derivative is zero or does not exist.

Linearization: Tangent-line approximation near an operating point.

Taylor polynomial: A finite derivative-based polynomial approximation about a chosen center.

Analytic function: A function locally equal to its convergent Taylor series.

Remainder: The difference between a function and its Taylor polynomial.

Definite integral: Signed accumulation over a specified interval.

Substitution: An integration technique based on reversing the chain rule.

Integration by parts: An integration technique derived from the product rule.

Improper integral: An integral defined through a limit because of an infinite interval or singularity.

Trapezoidal rule: A second-order numerical integration method using linear interpolation.

Simpson’s rule: A fourth-order numerical integration method based on quadratic interpolation.

Truncation error: Error caused by replacing an exact infinite or limiting process with a finite approximation.

Review Questions

Short Answer Questions

1. State the formal definition of convergence of a sequence.
2. List the three conditions for continuity at a point.
3. What units does dv/dt have when voltage is in volts and time is in seconds?
4. Why does infinite differentiability not guarantee equality with a Taylor series?
5. State the convergence condition for \int₁^\inftyx^-pdx.

Long Answer Questions

1. Explain the relationship among limits, continuity and differentiability, including relevant theorem conditions.
2. Derive the current response of an ideal RC charging circuit from its capacitor-voltage expression.
3. Develop Taylor’s theorem with the Lagrange remainder and explain how it provides an error bound.
4. Compare substitution, integration by parts and partial fractions, giving an appropriate example of each.
5. Derive the composite trapezoidal and Simpson formulas and discuss their convergence orders and limitations.

Analytical / Application Questions

1. A voltage is v(t)=12(1-e^-t/0.04)\ \mathrm{V}. Determine its initial rate of rise and its rate at t=0.04\ \mathrm{s}.
2. Approximate \ln(1.1) using the first four nonzero terms of its Maclaurin series and estimate the error using the alternating-series principle.
3. Current is modeled by i(t)=I₀e^-t/\tau, where I₀ is in amperes and \tau in seconds. Determine the charge transferred over [0,T] and its limiting value as T\to\infty.

Multiple Choice Questions with Answers

Question 1

Which statement correctly describes the sequence a_n=(1/3)ⁿ?

A. It converges to 0.

B. It converges to 1/3.

C. It diverges to infinity.

D. It oscillates without a limit.

Correct Answer: A. It converges to 0.

Question 2

Let f(x)=(x²-1)/(x-1) for x\ne1 and f(1)=k. Which value of k makes f continuous at 1?

A. 0

B. 1

C. 2

D. −1

Correct Answer: C. 2

Question 3

If voltage is measured in volts and time in seconds, what are the units of dv/dt?

A. V·s

B. V·s⁻¹

C. A·s

D. J·s⁻¹

Correct Answer: B. V·s⁻¹

Question 4

Which condition is sufficient for a function to equal its Taylor series in a neighborhood of the expansion point?

A. The function is continuous only.

B. The function has one derivative.

C. The function is infinitely differentiable, with no further condition.

D. The Taylor remainder tends to zero throughout that neighborhood.

Correct Answer: D. The Taylor remainder tends to zero throughout that neighborhood.

Question 5

The cubic Maclaurin approximation to e^0.2 is closest to which value?

A. 1.200000

B. 1.220000

C. 1.221333

D. 1.240000

Correct Answer: C. 1.221333

Question 6

Which substitution directly evaluates \int 2x\cos(x²)\,dx?

A. u=2x

B. u=x²

C. u=\cos x

D. u=1/x

Correct Answer: B. u=x²

Question 7

For which values of p does \int₁^\inftyx^-p\,dx converge?

A. p<0

B. p<1

C. p=1

D. p>1

Correct Answer: D. p>1

Question 8

Using the central difference with h=0.1, the estimate of d(\sin x)/dx at x=0.5 is approximately:

A. 0.389419

B. 0.564642

C. 0.876121

D. 1.000000

Correct Answer: C. 0.876121

Question 9

Under the required smoothness assumptions, the global truncation error of composite Simpson’s rule is of what order?

A. O(h)

B. O(h²)

C. O(h³)

D. O(h⁴)

Correct Answer: D. O(h⁴)

Question 10

For i(t)=kt² with k=4\ \mathrm{mA\,s^-2}, what charge is transferred from 0 to 2 s?

A. 8.000 mC

B. 10.667 mC

C. 12.000 mC

D. 16.000 mC

Correct Answer: B. 10.667 mC