Chebyshev Polynomial Series Overview

• Chebyshev polynomial series are expansions of functions using orthogonal Chebyshev polynomials, providing minimax optimal approximations on the interval [-1,1].
• Their inherent orthogonality facilitates efficient computation of interpolation, quadrature, and spectral methods, leading to rapid and accurate numerical solutions.
• The series offer exponential error decay for analytic functions and practical advantages in solving differential equations and other boundary value problems.

Chebyshev polynomial series are expansions of functions in terms of Chebyshev polynomials, central to approximation theory and numerical analysis on bounded intervals, particularly $[-1,1]$. Chebyshev polynomials provide a basis with optimal properties for minimizing maximal errors (minimax approximation), and their orthogonality facilitates efficient computation of interpolation, quadrature, and spectral methods for solving boundary value problems. Chebyshev polynomials are also foundational in harmonic analysis, waveform design, and partial differential equations, where their series expansions yield computationally superior alternatives to Fourier series for the numerical approximation of polynomial and analytic functions.

1. Definitions and Fundamental Properties

Let $T_n(x)$ denote the Chebyshev polynomials of the first kind, defined as

$T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$

yielding the explicit recurrence: $$T_0(x)=1, \quad T_1(x)=x, \quad T_{n+1}(x)=2xT_n(x) - T_{n-1}(x), \; n\geq 1.$$ The generating function for the Chebyshev polynomials is

$$G(x,t)=\sum_{n=0}^\infty T_n(x)\,t^n = \frac{1-xt}{1-2xt+t^2}.$$

Chebyshev polynomials are orthogonal with respect to the weight $w(x) = 1/\sqrt{1-x^2}$ on $[-1,1]$: $$\int_{-1}^1 T_m(x)T_n(x) w(x)\,dx = \begin{cases} 0 & \text{if } m\neq n,\ \pi & \text{if } m=n=0,\ \pi/2 & \text{if } m=n\geq1. \end{cases}$$ This orthogonality underpins the efficient computation of expansion coefficients and best-approximation polynomials (Karjanto, 2020).

2. Chebyshev Series Expansion and Coefficients

Any function $f$ that is piecewise continuous on $[-1,1]$ and integrable with $T_n(x)$0 can be expanded as

$T_n(x)$1

where the expansion coefficients are given by

$T_n(x)$2

Parseval’s identity in this context reads

$T_n(x)$3

providing a direct link between the $T_n(x)$4-norm (weighted) of $T_n(x)$5 and its expansion coefficients (Karjanto, 2020).

3. Comparison with Fourier Series and Approximation Quality

In the context of polynomial approximation, interpolation via Chebyshev polynomials is generally more accurate than using a Fourier series. Unlike the uniform spacing associated with Fourier and equispaced polynomial interpolation (prone to the Runge phenomenon), the use of Chebyshev nodes—roots or extrema of $T_n(x)$6—suppresses oscillations at the interval boundaries. Convergence properties are directly related to the analytic structure of $T_n(x)$7: if $T_n(x)$8 is analytic in an ellipse of the complex plane (the Bernstein ellipse) with foci $T_n(x)$9, there exist constants $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$0 such that

$T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$1

and the supremum-norm error decays as $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$2 upon truncation at degree $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$3. For functions with lower regularity, the coefficients decay algebraically, $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$4 for $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$5-fold continuously differentiable functions (Karjanto, 2020).

4. Differential Equations and Sturm-Liouville Connection

The Chebyshev differential equation, a form of the Sturm-Liouville problem,

$T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$6

admits Chebyshev polynomials $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$7 as its eigenfunctions. This connection is foundational for many boundary value problems encountered in mathematical physics, since the Sturm-Liouville form ensures real orthogonal eigenfunctions for self-adjoint operators. Spectral methods exploiting Chebyshev polynomial expansions are widely applied to numerically solve such problems with high accuracy and spectral convergence (Karjanto, 2020).

5. Example: Explicit Expansion and Process

For $T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$8, direct computation yields

$T_n(x) = \cos(n \arccos x), \quad |x| \le 1,$9

and all higher $$T_0(x)=1, \quad T_1(x)=x, \quad T_{n+1}(x)=2xT_n(x) - T_{n-1}(x), \; n\geq 1.$$0 vanish. Thus,

$$T_0(x)=1, \quad T_1(x)=x, \quad T_{n+1}(x)=2xT_n(x) - T_{n-1}(x), \; n\geq 1.$$1

This demonstrates that the Chebyshev expansion exactly recovers polynomials up to degree $$T_0(x)=1, \quad T_1(x)=x, \quad T_{n+1}(x)=2xT_n(x) - T_{n-1}(x), \; n\geq 1.$$2 when truncated at the appropriate order, highlighting the series’ practical efficiency for polynomial function approximation (Karjanto, 2020).

6. Applications and Advantages

Chebyshev polynomial series underpin numerous algorithms in computational mathematics:

• Optimal polynomial interpolation (Clenshaw–Curtis quadrature)
• Spectral and pseudospectral methods for PDEs
• Signal processing (minimax filter design)
• Efficient approximation of analytic functions and rapid summation
• Reduction of numerical errors relative to other series expansions (Karjanto, 2020)

Chebyshev approximation enables exponential convergence rates for analytic functions, produces near-best (minimax) uniform approximations, and—due to the discrete cosine transform connection—offers linear or nearly linear computational complexity for coefficient evaluations and transforms in numerical implementation.

In summary, the Chebyshev polynomial series forms the mathematical backbone for high-precision and robust polynomial approximation and spectral algorithms, with foundational advantages in accuracy, convergence, and computational tractability compared to traditional Fourier-based methods.