Skip to content
sssabryPublic

About

Taylor Model Enabled Reachability Analysis

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

9 Commits

Folders and files

Repository files navigation

Continuous Single Step Jet Engine

TERA

Taylor Model Enabled Reachability Analysis Framework.

TERA is a Python-native, open-source Taylor Model (TM) reachability framework for computing rigorous flowpipe enclosures of nonlinear dynamical systems under uncertainty. It unifies continuous, hybrid, and stochastic analysis within a single symbolic–numeric workflow to support rapid prototyping without sacrificing enclosure guarantees.


Highlights

  • Validated TM flowpipe computation for nonlinear ODEs
    • Local single-step TM integration (validated ODE solving in the sense of Berz/Makino-style TM integrators). Continuous Single Step Jet Engine
    • Left–right (compositional) propagation to mitigate wrapping over long horizons (“shrink-wrapping” preconditioning, formalized in later preconditioning work). Continuous Left-Right Quadratic Model
  • Hybrid reachability
    • Discrete transition processing with guard intersection and reset mapping, consistent with TM-based hybrid reachability semantics. Hybrid Bouncing Ball
  • Stochastic reachability
    • $\delta$-probabilistic reachable set ($\delta$-PRS) enclosures: sets that contain trajectories with probability at least 1 − $\delta$.
    • Computes $\delta$-PRS by combining deterministic TM flowpipes with probabilistic deviation bounds (AMGF-style separation strategy). Stochastic Spring Model
  • Visualization utilities
    • Plotting/projection tools for deterministic flowpipes and $\delta$-PRS enclosures, plus comparisons against Monte Carlo samples.
  • Examples and notebooks for continuous, hybrid, and stochastic workflows and benchmarks.

Project Layout

  • TERA/TMCore — core algebra: intervals, polynomials, Taylor models, TM vectors.
  • TERA/TMFlow — continuous reachability: flowpipe construction, remainder handling, preconditioning (single-step + left–right).
  • TERA/Hybrid — hybrid automata: modes, invariants, guard intersection, resets, worklist orchestration.
  • TERA/Stochastic — stochastic reachability and $\delta$-PRS workflows; sampling utilities for comparison.
  • TERA/Workbench — visualization and result helpers.
  • Examples — Jupyter notebooks for end-to-end usage.

Installation

See INSTALL.txt for up-to-date installation instructions. SageMath is required for core functionality, and installation is currently documented as an editable install inside a SageMath environment. Note: this project is not currently pip-installable; please use the SageMath environment workflow described in INSTALL.txt.

Examples/Benchmarks

Run one of the example notebooks in Examples/:

  • Continuous_*.ipynb

  • Hybrid_*.ipynb

  • Stochastic_*.ipynb

Configuration Notes

For full configuration details, see: Docs/Configuration_Ref.md and Docs/Workbench_API.md.

Status

Research prototype. APIs and configuration knobs may evolve as the engine matures. Current focus is consolidating hybrid + stochastic components into a unified SHS formulation. CI tests are planned for a future release (not yet enabled).

Notebook size: example notebooks in Examples/ include saved outputs. If repo size matters, clear notebook outputs before pushing.

License

This project is licensed under the GNU General Public License v2.0 (GPL-2.0). TERA is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License, version 2 (GPL-2.0). A copy of the license is provided in LICENSE.GPL2.

About

Taylor Model Enabled Reachability Analysis

Topics

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages