Computable Coordinate System Objects: Theory and Applications
Coordinate system transformations are fundamental in modern mathematics, physics, and engineering. Traditional approaches rely on matrix operations and tensor analysis: the former lacks geometric intuition, while the latter is overly abstract and challenging for numerical implementation. This paper proposes a specialized framework called Computable Coordinate Systems, which discretizes continuous transformations through Lie group-Lie algebra theory, achieving complete algebraic operability of coordinate systems. This paper systematically presents the comprehensive theoretical foundation of computable coordinate systems, including: (1) algebraic structure design based on group theory, defining operations such as multiplication, division, and exponential mapping; (2) Lie group-Lie algebra correspondence, establishing bridges between infinitesimal and finite transformations; (3) Berry connection and curvature theory, providing computable expressions for covariant derivatives and Riemann curvature tensors, correcting curvature formula errors in early research; (4) complete C++ implementation, including data structure design, operator overloading, and numerical algorithms. The numerical validation section demonstrates theoretical correctness through three typical application scenarios: (1) Computational Geometry surface operators, achieving <2% curvature error with O(h^2) convergence; (2) CAD topology verification, achieving 100% accuracy with <1% computational overhead; (3) Fourier spectral transformation, achieving <3.5% reconstruction error with O(h²) convergence rate. Experimental results show that computable coordinate systems significantly simplify implementation complexity while maintaining mathematical rigor, providing efficient and reliable computational tools for engineering applications.
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2025-10-22
- DOI
- https://doi.org/10.5281/zenodo.17413485
- Primary Topic
- Model Reduction and Neural Networks
- Type
- article
- Field-Weighted Citation Impact
- 0.00