Engineering Math Series 11: Matrices and Linear Systems

The starting point of linear algebra is not the matrix itself but the perspective of expressing and manipulating multiple equations at once. In engineering mathematics, a matrix is not a table of numbers but a language that compresses system structure.

Linear Systems Come First

Consider the following system of equations.

$$\begin{aligned} 2x + y &= 5 \\ x - y &= 1 \end{aligned}$$

Writing this in matrix form gives

$$\begin{pmatrix} 2 & 1 \\ 1 & -1 \end{pmatrix} \begin{pmatrix} x \\ y \end{pmatrix} = \begin{pmatrix} 5 \\ 1 \end{pmatrix}$$

That is, the form

$A\mathbf{x} = \mathbf{b}$

How to View Matrices

Matrix $A$ can be seen as a linear transformation that sends one vector to another. It is not just a table but a rule that stretches, shrinks, and rotates space.

With this perspective, three tasks all become unified under the same language:

• Solving systems of equations
• Changing coordinates
• Analyzing system structure

Worked Example

Let us directly solve the system above. From the second equation,

$x = 1 + y$

Substituting into the first equation,

$2(1+y) + y = 5$

gives

$2 + 3y = 5$

so

$y = 1, \quad x = 2$

From the matrix perspective, this solution means "the matrix $A$ sends the vector $(2,1)^T$ to $(5,1)^T$."

Why Linearity Is Important

When linearity holds, the superposition principle works. This is very important in engineering.

• The effects of multiple inputs can be combined
• The computational structure is regular
• It leads to more powerful tools like eigenvalues, diagonalization, and least squares

Nonlinear problems are more common in reality, but linear problems remain central tools due to their analyzability and designability.

Engineering Applications

Circuit Network Analysis

Using Kirchhoff's laws, the relationships among multiple currents and voltages are organized as systems of linear equations.

Structural Analysis

Under the small deformation assumption, force-displacement relationships are modeled as linear systems.

Machine Learning

Linear regression is ultimately a problem of solving matrix equations, and data processing mostly runs on matrix operations.

When Solutions Do Not Exist or Are Not Unique

Not every linear system has a unique solution.

• No solution: contradictory equations
• Infinitely many solutions: equations are not independent
• Unique solution: the matrix provides sufficient information

This distinction becomes clearer when we learn about determinants and rank later.

Common Mistakes

Viewing matrices as mere computation objects

Matrices carry structure. You must read what each column and row means.

Mixing up the variable order

In $A\mathbf{x}=\mathbf{b}$, changing the vector order changes the meaning of the entire equation.

Thinking of substitution and matrix perspectives separately

Even if you use substitution or elimination by hand, it is good not to lose sight of the fact that the underlying structure is the same linear system.

One-Line Summary

Matrices are the common language of engineering mathematics for expressing and analyzing multiple linear relationships at once.

Next Post Preview

In the next post, we will organize the core tools for actually solving linear systems: elimination, inverse matrices, and determinants.

References

• Erwin Kreyszig, Advanced Engineering Mathematics, 10th Edition
• Gilbert Strang, Introduction to Linear Algebra
• David C. Lay, Linear Algebra and Its Applications