Mathematical and Algebraic Foundations

Introduction

Dynamic systems in mechatronics are commonly represented by differential equations, difference equations, transfer functions, and state-space models. All of these representations depend on a small set of mathematical ideas: numbers may be complex, variables may be related through functions, approximations may be expressed as series, and multivariable relationships may be organized using matrices and vectors. Eigenvalues and eigenvectors then reveal important system properties such as natural modes, stability, resonance, and decoupling.

This chapter reviews these foundations with emphasis on interpretation and engineering use. The focus is not merely computational. A mechatronics engineer must also know when a mathematical operation is valid, what assumptions it makes, and how its result relates to a physical system.

Learning Objectives

By the end of this chapter, you should be able to:

1. Represent and manipulate complex numbers in Cartesian, polar, and exponential form.
2. Use Euler’s formula to connect exponential, trigonometric, and sinusoidal descriptions.
3. Classify functions and explain limits, continuity, and composition.
4. Use power and Taylor series to approximate functions and assess convergence.
5. Perform matrix operations and calculate determinants and inverses.
6. Explain vector spaces, subspaces, span, basis, dimension, and linear independence.
7. Determine matrix rank and interpret it in terms of independent information.
8. Calculate and interpret eigenvalues and eigenvectors.
9. Diagonalize suitable matrices and identify cases in which diagonalization fails.
10. Apply these concepts to a dynamic mechatronic system.

1. Complex Numbers

The real-number system is insufficient for solving some polynomial equations. For example, the equation x² + 1 = 0 has no real solution. The imaginary unit is defined by

A complex number has the form

where a and b are real numbers. The real part is Re(z) = a, and the imaginary part is Im(z) = b. Two complex numbers are equal only when their real and imaginary parts are separately equal.

Arithmetic

For z₁ = a + bi and z₂ = c + di,

The complex conjugate of z = a + bi is z^* = a - bi. Since zz^* = a² + b², division can be performed by multiplying numerator and denominator by the conjugate:

The modulus, or magnitude, is

It is a nonnegative real number and may be interpreted geometrically as the distance from the origin to the point (a,b) in the complex plane.

Polar and exponential forms

If z = a + bi and θ is the argument measured from the positive real axis, then

where r = |z|. Euler’s formula states that

Thus,

The angle must be selected with the correct quadrant, commonly using θ = atan2(b,a). Polar form makes multiplication and division especially simple:

De Moivre’s theorem follows directly:

for integer n.

Worked Example 1: Complex representation of a sinusoid

Consider z = 3 - 4i. Its magnitude is

The argument is θ = atan2(-4,3) ≈ -0.927 rad. Therefore,

In electrical and control engineering, a sinusoidal signal may be represented by the real part of a complex exponential:

where X is a complex amplitude and ω is angular frequency in rad/s. This is a computational representation; the physical signal remains real.

2. Functions and Their Engineering Interpretation

A function assigns each permitted input exactly one output. In notation,

The set of permitted inputs is the domain, and the set of possible outputs is the range. In dynamic systems, functions may map time to displacement, input voltage to motor torque, or a vector of states to a derivative vector.

Important categories include:

• A scalar function maps scalar inputs to scalar outputs.
• A vector-valued function maps a scalar or vector input to a vector output.
• A linear function satisfies f(αx + βy) = αf(x) + βf(y) for all permitted scalars α,β and vectors x,y.
• A nonlinear function violates this superposition property.
• A time-invariant relation does not change when the time origin is shifted.

For example, f(x) = 3x - 2 is affine but not linear because f(0) = -2 ≠ 0. The function f(x) = 3x is linear.

Composition and inverse functions

If g maps x to u and f maps u to y, their composition is

An inverse function f^-1 exists on a specified domain when f is one-to-one. Then f^-1(f(x)) = x. Care is required: a function may have an inverse only after its domain is restricted.

Limits and continuity

The statement lim_x→a f(x) = L means that values of f(x) approach L as x approaches a. A function is continuous at a when

Continuity is important in physical modeling because abrupt mathematical discontinuities may represent switching, impacts, saturation, or idealized control actions. Such features should be identified rather than inadvertently introduced.

3. Sequences and Series

A sequence is an ordered list {a_n}, while a series is the sum of sequence terms:

A series converges if its partial sums approach a finite limit. Convergence is essential when an infinite expansion is used to approximate a physical or computational quantity.

Common series

The geometric series satisfies

The condition |r| < 1 is essential. If |r| ≥ 1, the series does not converge to the stated value.

The exponential, sine, and cosine functions have Maclaurin series:

These series converge for every real x. For a general function, the Taylor expansion about x = a is

provided the function is sufficiently differentiable and the series converges to the function in the region being used. Truncation produces an approximation, not an identity.

Worked Example 2: Local approximation

For small x,

At x = 0.2 rad, this gives cos(0.2) ≈ 1 - 0.04/2 = 0.98. The actual value is approximately 0.98007, so the absolute error is about 0.00007. The approximation is effective because the omitted fourth-order term is small. The angle must be in radians for these series formulas.

In engineering, series expansions support linearization, numerical algorithms, sensor calibration, and approximations of nonlinear models near an operating point.

4. Matrices and Determinants

A matrix is a rectangular arrangement of numbers or symbols. An m × n matrix has m rows and n columns:

A vector is often treated as a one-column matrix. Matrix notation makes coupled equations compact. For example,

represents two simultaneous equations.

Operations

Addition requires equal dimensions and is element-by-element. Multiplication AB is defined when the number of columns of A equals the number of rows of B. Its entries are

In general, AB ≠ BA. The identity matrix I satisfies AI = IA = A for compatible dimensions. The transpose is A^T, obtained by interchanging rows and columns.

Determinants and inverses

For a 2 × 2 matrix,

A square matrix is invertible if and only if its determinant is nonzero. For the same 2 × 2 matrix,

provided ad - bc ≠ 0. For larger matrices, determinants may be computed by cofactor expansion or elimination, although numerical software generally uses more stable factorization methods.

A system Ax = b has a unique solution when A is square and nonsingular. If A is singular, the system may have no solution or infinitely many solutions, depending on b.

Worked Example 3: Solving a coupled system

Let

The determinant is 2×3 - 1×1 = 5, so the matrix is invertible. Its inverse is

Therefore,

Substitution gives 2(1.6)+1.8=5 and 1.6+3(1.8)=7, confirming the result.

5. Vector Spaces, Span, and Linear Independence

A vector space is a set closed under vector addition and scalar multiplication and satisfying the standard algebraic rules for these operations. Examples include ℝⁿ, polynomial spaces, and spaces of continuous functions.

The span of vectors v₁,…,v_k is the set of all linear combinations:

The vectors are linearly independent if

implies c₁ = ⋯ = c_k = 0. If a nonzero combination produces the zero vector, the vectors are dependent.

A basis is a linearly independent spanning set. The number of vectors in any basis is the dimension of the vector space. In state-space modeling, a basis corresponds to a choice of state coordinates. Different bases may describe the same physical system.

A subspace is a subset that is itself a vector space. The null space of A is

and the column space is the span of the columns of A.

6. Rank and Gaussian Elimination

The rank of a matrix is the dimension of its column space; equivalently, it is the maximum number of linearly independent rows or columns. Row reduction to echelon form reveals the rank through the number of pivots.

For an m × n matrix,

The rank-nullity theorem states

where n is the number of columns and nullity is the dimension of the null space.

Rank has practical meaning. It indicates how many independent directions or constraints are present. In a sensor system, deficient rank may mean that measurements cannot distinguish all states. In a mechanism, it may indicate a kinematic constraint or singular configuration.

For a square matrix, full rank is equivalent to invertibility. For a rectangular system Ax=b, consistency requires that b belong to the column space of A.

7. Eigenvalues and Eigenvectors

For a square matrix A, a nonzero vector v is an eigenvector if

where λ is the corresponding eigenvalue. Rearranging gives

A nonzero solution exists only when

This is the characteristic equation. Its roots are the eigenvalues. Eigenvalues may be real or complex, even when A has only real entries. For a real matrix, nonreal eigenvalues occur in conjugate pairs.

Worked Example 4: Eigenanalysis

Let

The characteristic equation is

Thus λ₁ = 3 and λ₂ = 1. For λ₁=3,

which gives an eigenvector proportional to v₁=[1;1]. For λ₂=1, an eigenvector is v₂=[1;-1]. Direct multiplication verifies Av₁=3v₁ and Av₂=v₂.

In continuous-time linear systems ẋ = Ax, eigenvalues describe modal behavior e^λt. A mode with negative real part decays, one with positive real part grows, and one with zero real part requires further analysis. In discrete-time systems x_k+1=Ax_k, stability is associated with eigenvalues inside the unit circle, |λ|<1, under the usual asymptotic-stability conditions.

8. Diagonalization

A matrix is diagonalizable if there is an invertible matrix P such that

where D is diagonal. The columns of P are eigenvectors, and the corresponding diagonal entries of D are eigenvalues. Equivalently,

Diagonalization is useful because powers and exponentials become simpler:

and, when defined,

For diagonal D = diag(λ₁,…,λ_n), D^k = diag(λ₁^k,…,λ_n^k).

An n × n matrix with n linearly independent eigenvectors is diagonalizable. Distinct eigenvalues guarantee independent eigenvectors, but repeated eigenvalues do not guarantee them. For example,

has only the eigenvalue λ=1 and only one independent eigenvector direction, so it is not diagonalizable. Such matrices require Jordan-form methods or direct matrix-exponential techniques.

9. Applied Case: A Two-State Mechatronic Model

Background

Consider an illustrative linearized model of a coupled positioning mechanism with state vector x=[x₁;x₂]. The first state may represent a normalized displacement and the second a normalized velocity or an interacting actuator state. Around an operating point, suppose the discrete-time model is

The entries are dimensionless transition coefficients for a chosen sampling interval.

Problem

Determine whether the unforced model is asymptotically stable and identify its dominant mode.

Analysis

The characteristic equation is

Expanding,

The roots are λ₁=0.9 and λ₂=0.6. Both satisfy |λ|<1, so the model is asymptotically stable in discrete time. The eigenvalue 0.9 is closest to the unit circle and therefore represents the slow, dominant decay mode.

For λ₁=0.9, an eigenvector satisfies

so one choice is v₁=[2;1]. For λ₂=0.6, one choice is v₂=[1;-1]. With

we have A=PDP^-1. In modal coordinates z=P^-1x, the system becomes z_k+1=Dz_k, so the two coupled states are represented as two independent modal sequences.

Outcome and engineering interpretation

Any initial condition can be decomposed into the two eigenvector directions. The component along v₁ decays as 0.9^k, while the component along v₂ decays faster as 0.6^k. This analysis helps estimate settling behavior and identify which mode a controller must suppress most strongly.

This is an illustrative case, not a complete physical design. A real model would require units, sampling-period information, actuator limits, sensor noise, parameter uncertainty, and validation against experimental data.

10. Limitations, Numerical Practice, and Professional Use

Exact symbolic calculations are valuable for understanding, but engineering models often contain measured parameters and floating-point numbers. Near-singular matrices can produce large numerical errors. A small determinant alone is not a complete measure of numerical reliability; condition numbers and appropriate matrix factorizations should also be considered.

Best practice includes:

• Use radians in calculus, Taylor expansions, and angular-frequency formulas.
• Check matrix dimensions before multiplying.
• Avoid explicitly computing A^-1 when solving Ax=b numerically; use a reliable linear solver or factorization.
• Verify eigenpairs by evaluating the residual Av-λv.
• Preserve units when constructing state variables and system matrices.
• Distinguish a mathematical model from the physical system it approximates.
• Report assumptions, operating point, sampling time, and parameter sources.
• Use software as a verification tool, not as a substitute for interpretation.

In safety-critical mechatronic systems, incorrect eigenvalue or rank conclusions can lead to unstable controllers or undetected sensor deficiencies. Independent checks, simulation, testing, and documented review are therefore part of responsible engineering practice.

Chapter Summary

Complex numbers extend the real numbers and provide a compact representation of magnitude and phase. Euler’s formula connects complex exponentials with sine and cosine, enabling efficient sinusoidal analysis. Functions describe mappings between variables, while series provide controlled approximations when convergence and truncation error are understood.

Matrices organize coupled equations and transformations. Determinants characterize invertibility for square matrices, and rank measures independent rows, columns, constraints, or information. Vector spaces provide the language of span, basis, dimension, and linear independence. Eigenvalues and eigenvectors identify special directions in which a linear transformation acts by simple scaling. If enough independent eigenvectors exist, diagonalization transforms a coupled model into independent modal coordinates.

These ideas form the algebraic foundation of state-space modeling, stability analysis, numerical simulation, robotics, signal processing, control engineering, and dynamic-system design.

Key Terms

Complex number: A number of the form a+bi, where a,b are real.

Complex conjugate: The number obtained by changing a+bi to a-bi.

Modulus: The magnitude |z| of a complex number.

Euler’s formula: The identity e^iθ=cosθ+i sinθ.

Function: A mapping that assigns each permitted input exactly one output.

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

Sequence: An ordered list of terms indexed by an integer.

Series: The sum of terms in a sequence.

Taylor series: A derivative-based expansion of a function about a point.

Matrix: A rectangular array of numbers or symbols.

Identity matrix: A square matrix that leaves compatible vectors unchanged.

Determinant: A scalar associated with a square matrix that indicates, among other properties, invertibility.

Vector space: A set closed under vector addition and scalar multiplication.

Basis: A linearly independent spanning set.

Rank: The dimension of a matrix’s column space.

Null space: The set of vectors mapped to zero by a matrix.

Eigenvalue: A scalar satisfying Av=λv for a nonzero vector v.

Eigenvector: A nonzero vector whose direction is unchanged by a matrix transformation.

Diagonalization: Representation of a matrix as A=PDP^-1.

Modal coordinate: A coordinate associated with an eigenvector and its system mode.

Review Questions

Short Answer Questions

1. Define the modulus and complex conjugate of z=a+bi.
2. State Euler’s formula and explain why it is useful in engineering.
3. What condition makes a square matrix invertible?
4. Define linear independence in terms of a linear combination.
5. What does matrix rank measure?

Long Answer Questions

1. Explain Cartesian, polar, and exponential forms of a complex number and derive the multiplication rule in polar form.
2. Explain Taylor-series approximation, including the importance of convergence, truncation, and radians.
3. Describe the relationship among column space, null space, rank, and nullity.
4. Derive the characteristic equation for a 2 × 2 matrix and explain how its roots are used.
5. Explain diagonalization and discuss why repeated eigenvalues may prevent diagonalization.

Analytical / Application Questions

1. For A=[4 1; 2 3], calculate the eigenvalues and determine whether the matrix is diagonalizable.
2. A discrete-time model has eigenvalues 0.98, 0.7, and 1.02. Assess asymptotic stability and identify the mode requiring engineering attention.
3. A sensor matrix has rank 2 but three columns. Use rank-nullity to determine the nullity and explain what the result could mean for state reconstruction.

Multiple Choice Questions with Answers

Question 1

Which expression is the complex conjugate of 3-4i?

A. 3+4i

B. -3+4i

C. -3-4i

D. 4-3i

Correct Answer: A. 3+4i

Question 2

What is the polar magnitude of 1+i?

A. 1

B. √2

C. 2

D. 0

Correct Answer: B. √2

Question 3

Euler’s formula states that e^iθ equals:

A. sinθ+i cosθ

B. cosθ-i sinθ

C. cosθ+i sinθ

D. e^θ(cosθ+sinθ)

Correct Answer: C. cosθ+i sinθ

Question 4

For the geometric series ∑_n=0^∞rⁿ, which condition guarantees convergence?

A. r>1

B. |r|<1

C. r=1

D. |r|=2

Correct Answer: B. |r|<1

Question 5

What is the determinant of [2 1; 3 4]?

A. 5

B. 8

C. 11

D. −5

Correct Answer: A. 5

Question 6

A set of vectors is linearly independent when:

A. Every vector has unit length

B. Their span contains only the zero vector

C. Only the trivial linear combination equals the zero vector

D. All vectors are mutually perpendicular

Correct Answer: C. Only the trivial linear combination equals the zero vector

Question 7

If a matrix has three columns and rank two, what is its nullity?

A. 1

B. 2

C. 3

D. 5

Correct Answer: A. 1

Question 8

The eigenvalues of a matrix satisfy which equation?

A. det(A+λI)=1

B. det(A-λI)=0

C. Aλ=0

D. trace(A-λI)=0

Correct Answer: B. det(A-λI)=0

Question 9

A discrete-time linear system is asymptotically stable when, under standard assumptions, its eigenvalues:

A. All have positive real parts

B. All are greater than one

C. All lie strictly inside the unit circle

D. All are equal to zero

Correct Answer: C. All lie strictly inside the unit circle

Question 10

A square matrix with a complete set of linearly independent eigenvectors can be:

A. Transposed only

B. Diagonalized

C. Made singular by definition

D. Converted only to a zero matrix

Correct Answer: B. Diagonalized