pyncbt 0.1.13

Non-iterative Correlation-based Tuning (NCbT) for data-driven control

pyncbt · Non-iterative Correlation-based Tuning (NCbT) in Python

pyncbt is an open‑source Python library that implements Non‑iterative Correlation‑based Tuning (NCbT) for direct fixed‑structure controller design from input‑output data. It requires no explicit plant model and no iterative optimization.

• Simple workflow: weighting filter W → instruments → stable least‑squares → parameters ρ.
• Reproducible examples: .npy data, benchmark scripts, and Colab notebook.
• Open‑loop and closed‑loop support: flexible for different experimental scenarios.
• Flexible basis: you define the controller structure (PID, FIR, integrator plus delays, etc.).

Full documentation: https://joaoleroy.github.io/pyncbt/
PyPI package: https://pypi.org/project/pyncbt/

---

Installation

pip install pyncbt

---

Interactive notebooks

Try the examples online without any installation:

[https://colab.research.google.com/drive/1MlzWRP8eHVUVVPgt1IiL2hqU5SN7gq98?usp=sharing] (Benchmark example (flexible transmission system))

[https://colab.research.google.com/drive/1zblNhjnmtO1QX1wz6xGg9ekQapmgzl7u?usp=sharing] (PI controller for a non‑minimum phase system)