Variable-Order Springpot Model Tuning via Cross-Entropy Method

SpringpotTune: Variable-Order Springpot Model Tuning via Cross-Entropy Method is a Matlab package designed to fit variable-order springpot models (fractional rheological elements) using the Cross-Entropy (CE) method for global optimization. With simple adaptations, the SpringpotFit framework can be extended to optimize other variable-order fractional models in rheology and time-dependent systems, offering a versatile and robust approach to parameter identification.

SpringpotTune uses as optimization tool the package CEopt - Cross-Entropy Optimizer, which can be downloaded at https://ceopt.org.

Table of Contents

• Overview
• Features
• Usage
• Documentation
• Reproducibility
• Authors
• Citing FracTune
• License
• Institutional support
• Funding

Overview

SpringpotTune was developed to fit variable-order springpot models for rheological systems by addressing nonconvex optimization problems using the Cross-Entropy (CE) method. The underlying results are reported in the following publication:

• J. G. Telles Ribeiro and A. Cunha Jr, Advanced creep modeling of polypropylene: A variable-order fractional calculus approach, 2024 (under review)

Preprint available here.

Features

• Implements the Cross-Entropy method for fitting variable-order springpot models
• Transparent "gray-box" optimizer offering intuitive control over key parameters
• Robust and scalable for moderately complex fitting problems in rheological systems
• Demonstrated versatility and effectiveness through case studies of variable-order fractional models

Usage

To get started with SpringpotTune, follow these steps:

1. Clone the repository:

   git clone https://github.com/americocunhajr/SpringpotTune.git

2. Navigate to the code directory:

   cd SpringpotTune/SpringpotTune-1.0

3. To calibrate the Springpot model, execute the main file corresponding to your case:

   Main_ModelCalibration_VOSpringpot

This package includes the following files:

• CEopt.m -- Cross-entropy solver
• Main_ModelCalibration_VOSpringpot.m -- Model calibration for the variable-order springpot
• PlotSemilogx2.m -- Auxiliary function to plot two temporal curves from given datasets in semilogx scale

Documentation

The routines in SpringpotTune are well-commented to explain their functionality. Each routine includes a description of its purpose, inputs, and outputs.

Reproducibility

Simulations done with SpringpotFit are fully reproducible, as can be seen on this CodeOcean capsule.

Authors

• José Geraldo Telles Ribeiro
• Americo Cunha Jr

Citing SpringpotTune

We ask the code users to cite the following manuscript in any publications reporting work done with our code:

• J. G. Telles Ribeiro and A. Cunha Jr, Advanced creep modeling of polypropylene: A variable-order fractional calculus approach, 2024 (under review)

@article{TellesRibeiro2024SpringpotTune,
   author  = {J. G. {Telles Ribeiro} and A. {Cunha~Jr}},
   title   = "{Advanced creep modeling of polypropylene: A variable-order fractional calculus approach}",
   journal = {Under Review},
   year    = {2024},
   volume  = {~},
   pages   = {~},
   doi    = {~},
}

License

SpringpotTune is released under the MIT license. See the LICENSE file for details. All new contributions must be made under the MIT license.

Institutional support

Funding