Comparison of the methods for discrete approximation of the fractional-order operator

L. Dorcak,I. Petras,J. Terpak,M. Zborovjan

DOI: https://doi.org/10.48550/arXiv.math/0306017

2003-06-01

Abstract:In this paper we will present some alternative types of discretization methods (discrete approximation) for the fractional-order (FO) differentiator and their application to the FO dynamical system described by the FO differential equation (FDE). With analytical solution and numerical solution by power series expansion (PSE) method are compared two effective methods - the Muir expansion of the Tustin operator and continued fraction expansion method (CFE) with the Tustin operator and the Al-Alaoui operator. Except detailed mathematical description presented are also simulation results. From the Bode plots of the FO differentiator and FDE and from the solution in the time domain we can see, that the CFE is a more effective method according to the PSE method, but there are some restrictions for the choice of the time step. The Muir expansion is almost unusable.

Optimization and Control

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: to compare the effectiveness of different discretization methods for discrete approximation of fractional - order operators, especially the application of these methods in fractional - order differential equations (FDE). Specifically, the author evaluates several effective discretization methods through the comparison of analytical solutions and numerical solutions, including the Tustin operator with Muir expansion and the continued fraction expansion (CFE) combined with the Tustin operator and the Al - Alaoui operator.

### Research Background and Problem Description

1. **Research Background**:
- Fractional - order calculus (FOCA) has a history of about 300 years, and its theory was mainly developed in the 19th century.
- In recent decades, the research on fractional - order derivatives and integrals is not only limited to the theoretical level, but also has been widely applied in many fields such as long power lines, electrochemical processes, dielectric polarization, colored noise, viscoelastic materials, chaotic phenomena, control theory, etc.

2. **Research Problems**:
- The paper aims to explore the approximation effects of different discretization methods on fractional - order differentiators (FO differentiator) or integrators (FO integrator).
- It mainly compares two effective methods: the Tustin operator with Muir expansion and the continued fraction expansion method (CFE), and compares them with the power series expansion method (PSE).
- The research also includes the application of these methods in fractional - order dynamic systems, especially evaluating their performance through Bode diagrams and time - domain responses.

### Main Research Contents

1. **Discretization Methods**:
- **Direct Discretization**: Directly discretize the continuous - time system.
- **Indirect Discretization**: First fit the transfer function of the continuous - time system in the frequency domain, and then discretize it.

2. **Specific Methods**:
- **Power Series Expansion Method (PSE)**: An approximate method based on power series expansion.
- **Continued Fraction Expansion Method (CFE)**: Continued fraction expansion combined with the Tustin operator or the Al - Alaoui operator.
- **Muir Expansion**: A recursive method applied to the Tustin generating function.

3. **Performance Evaluation**:
- Compare the performance of different methods in fractional - order differentiators and FDE through Bode diagrams and time - domain responses.
- The results show that the CFE method is more effective than the PSE method, but there are some limitations in choosing the time step. The Muir expansion is almost unusable.

### Key Formulas

- **Approximation of Fractional - order Differentiator**:
\[
D^\delta(s)=T^\delta s^\delta
\]
where \( T \) is the sampling period, and \( \delta \) is the order of the differentiator or integrator.

- **Fractional - order Differential Equation**:
\[
y(t)+a_1 D^\delta y(t)=a_0 u(t)
\]
where \( \delta \) is the differential order, and \( a_1 \) and \( a_0 \) are arbitrary constants.

- **Bode Characteristic Analysis**:
\[
FD(s)=T^\delta s^\delta
\]
\[
F(s)=\frac{1}{a_1 s^\delta - a_0}
\]

### Conclusion

Through a detailed comparison of different discretization methods, the author concludes that the continued fraction expansion method (CFE) is superior to the power series expansion method (PSE) in most cases, but the time step needs to be carefully selected to ensure stability. The Muir expansion is almost unusable.